Report No. 23119 The Gender Dimension of Bank …documents.worldbank.org/curated/en/945421468781751800/...The Gender Dimension of Bank Assistance: An Evaluation of Results This evaluation

Report No. 23119

The Gender Dimension of Bank AssistanceAn Evaluation of Results

January 17, 2002

Operations Evaluation Department

Document of the World Bank

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Abbreviations and Acronyms

ADB Asian Development BankBA Beneficiary AssessmentBP Bank ProcedureCAE Country Assistance EvaluationCAS Country Assistance StrategyCDF Comprehensive Development FrameworkCEDAW Convention on the Elimination of All Forms of Discrimination against WomenCHC Community Health CentersDAC Development Assistance CommitteeDFID Department for International DevelopmentEAP Environmental Action PlanESW Economic and sector workFAO Food and Agriculture OrganizationFP/MCH Family Planning/Maternal and Child HealthFY Fiscal yearGAD Gender and DevelopmentGDI Gender Related Development IndexGDP Gross domestic productGEM Gender Empowerment MeasureGP Good PracticeGSB Gender Sector BoardHNP Health, Nutrition and PopulationIADB Inter-American Development BankICR Implementation Completion ReportIDA Intemational Development AssociationIDF Institutional Development FundIEC Information, Education, and CommunicationIEPS Initial Executive Project SumrnaryISNAR Intemational Service for National Agricultural ResearchLIL Learning and Innovative LoanLSMS Living Standards Measurement SurveysMOE Ministry of EducationMOP Memorandum of the PresidentNGO Nongovernmental organizationOD Operational DirectiveOECD Organization for Economic Cooperation and DevelopmentOED Operations Evaluation DepartmentOEDCR Operations Evaluation Department Country Evaluation and Regional Relations GroupOMS Operational Manual StatementOP Operational PolicyPA Poverty AssessmentPAD Project Appraisal DocumentPCD Project Concept DocumentPER Public Expenditure ReviewPIF Project Information FormPREM Poverty Reduction and Economic ManagementPSR Project Status ReportQAG Quality Assurance GroupSAP Structural Adjustment ProgramsSAR Staff Appraisal ReportSSP Sector Strategy PapersTTL Task Team LeaderUNDP United Nations Development ProgramUNRISD United Nations Research Institute for Social DevelopmentUSAID United States Agency for International DevelopmentWID Women in Development

Director General, Operations Evaluation Department: Robert PicciottoDirector, Operations Evaluation Department: Gregory K. IngramManager, Country Evaluation and Regional Relations: Ruben LamdanyTask Manager, Country Evaluation and Regional Relations: Gita Gopal

The World BankWashington, D.C. 20433

U.S.A.

Office of the Director-GeneralOperations Evaluation

January 17, 2002

MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT

The Gender Dimension of Bank Assistance: An Evaluation of Results

This evaluation examines the results of the Bank's Women in Development/Gender andDevelopment policy over the last decade. It addresses the following questions: (i) to what extent didthe Bank help to reduce gender disparities in the health and education sectors? (ii) to what extent didthe Bank increase the participation of women in economic activities? and (iii) to what extent did theBank influence institutional changes that support the advancement of women? This report wasdiscussed by the Committee on Development Effectiveness on May 9, 2001, and a report of thatdiscussion is attached as Annex XI.

The evaluation assesses Bank assistance in twelve countries (Bangladesh, Cote d'Ivoire,Ecuador, Gambia, Haiti, Kyrgyz Republic, Philippines, Poland, Sri Lanka, Vietnam, Yemen andZambia) with varying degrees of gender disparity. It finds that Bank assistance has achievedsatisfactory results in the health and education sectors. The Bank has, however, been weak inpromoting the economic participation of women and in improving the Borrower's institutionalframework for gender, thereby reducing the overall development effectiveness of its assistance at acountry level.

The Bank has positively influenced the advancement of women in Bangladesh and TheGambia. Although causality is difficult to establish conclusively, the evaluation finds the followingkey common factors in Bank assistance to these countries:(i) country ownership and governmentcommitment as reflected by active policies and action plans for women and by strong andaccountable counterparts; (ii) Bank assistance consistent with the country's action plan; (iii) soundanalytical work as foundation for Bank operations; (iv) a holistic approach to gender issues thataddresses gender issues in the education and health sectors, aims to increase the participation ofwomen in the economic sectors and deals with key institutional aspects related to gender; and (v)effective collaboration with local and international partners. Bank programs in the other ten countrieswere not characterized by the above factors and did not produce similarly positive results.

The attached report reiterates the recommendation of OED Gender Evaluation 2000 that theBank's new strategy should clarify how the Bank's gender policy links with its poverty reductionmandate. It also offers the following recommendations: (i) strengthen Borrower institutions tosupport the implementation of national gender policies and/or action plans; in countries where suchpolicies are weak, support for their strengthening (through country dialogue and non-lendingservices) should be a Bank priority; (ii) integrate gender considerations into Country AssistanceStrategies based on a comprehensive diagnosis and through activities explicitly related to theBorrower's policy framework for gender; (iii) integrate gender into the design of Bank-supportedprojects so that both men and women are able to access the benefits equitably, especially in countrieswith high gender disparities.

2

A Gender Mainstreamning Sector Strategy Paper (GSSP) was endorsed by the Executive Boardon September 18, 2001. Copies of this report will be available from the Bank Public InformationCenter, or through the World Bank Web Site (the Executive Summary of the GSSP is attached asAnnex XIII of this report). The GSSP addresses the main issues highlighted in this and the previousOED evaluation reports (see Management Action Record in Attachment 2). The strategy involvesworking closely with countries to prepare periodic, multisectoral Country Gender Assessments thatanalyze the gender dimensions of development; identifying gender-responsive policies and actionsimportant for development effectiveness in the country; and integrating these policies and actions intopolicy dialogue and the country assistance programs. It establishes accountability for gendermainstreaming with the country director, requires the establishment of results based monitoringsystems, and asks for periodic reports by operational Vice-Presidents on progress in policyimplementation.

Acknowledgements

The gender team of the Country Evaluation and Regional Relations group of the OperationsEvaluation Department (OED), Ananya Basu, Diana Campillo, Anna Maria Inverso, Sarah Lubega,Inmaculada Montero-Luque, Vera Paquete Perdigao, Brandie Sasser, and Svenja Weber-Venghausresearched and contributed significantly to the preparation of this evaluation report. The team is alsograteful to Roziah Baba for her dedicated and strong support throughout the evaluation.

Comments from: Rema Balasundaram, Shamit Chakravarti, Poonam Gupta, Catherine Gwin, PatrickGrasso, John H. Johnson, Luis Landau, Elizabeth Morris Hughes, S. Ramachandran, and Cornelievan Waegeningh are gratefully acknowledged.

The OEDCR team also acknowledges the guidance and input of the external review group-including Mayra Buvinic, Geeta Rao Gupta, Jean King and Linda Morra. It acknowledges thesupport and guidance of the Gender and Development Board management and staff (Karen Mason,Cecilia Valdivieso, Adyline Waafas Ofosu-Amaah, Michael J. Bamberger, Lucia Fort) and otherregional gender coordinators (Mark Blackden, Maria Correia, Shaha Raza, Karin Kapadia, andGillian Brown) for their extensive support and collaboration through the various stages of thisevaluation. It also gratefully acknowledges the contribution of the participants of the brainstormingsession organized to initiate this evaluation.

The team also thanks the numerous participants and co-organizers (United Nations DevelopmentProgramme, United Nations Economic Commission for Europe, International Service for NationalAgricultural Research (ISNAR), United Nations Development Fund for Women (UNIFEM), andInter-American Development Bank) at the workshops in Nairobi, Quito and Warsaw. The team isalso grateful to the Evaluations Department of the Asian Development Bank, in particular HemmalaHettige, for their contribution and collaboration.

OED is grateful to the Swiss Development Corporation and the Norwegian Government for theirsupport in conducting the regional workshops and the beneficiary assessments, without which thisevaluation would have been very difficult. The team also thanks Meera Shah who organized thebeneficiary assessments (India-Uttar Pradesh with Praxis, Institute for Participatory Practices, SriLanka with the Institute for Participatory Interaction in Development and Vietnam), and the variouslocal teams, in particular Charity Kabutha, John Milimo (Zambia), Jurgette Honculada, VeronicaFenix Villavicencio (Philippines), Alexandra Martinez, Jorge Corral, Maria Moreno, Judith Segarra(Ecuador), and Dr. Vu Manh Loi, Le Thi Quy (Vietnam), who managed the country specificassessments. OED specially acknowledges the contribution of Ellen Goldstein (OEDCR) and HelenHambly Odame (ISNAR) in developing the methodology for and guiding the preparation of theZambia beneficiary assessments.

OED appreciates the contributions of the country teams of the 12 evaluation countries. In particular,the evaluation would like to thank Andrew Steer, Mahmoud Ayub, Frederick Temple, Isabel Grasso,Vinay Bhargava, Basil Kavalsky, and Laurence Clark, for their support in conducting in-countryevaluations. We would also like to specially acknowledge the contribution of the innumerablestakeholders and project beneficiaries who provided the building blocks for this evaluation.

In particular, the team would like to acknowledge the participation of Jan Piercy in all the regionalgender workshops. Her inspiring presence helped to elevate the discussions at these workshops.

CONTENTS

TERMS AND DEFINITIONS ...................................................................... i......... ......... ...i

1. The Evaluation Context. 1

2. Relevance Of Bank Assistance .3Soundness of Analytical Work .......... 4.................................... 4

Integrating gender into analytical work. 4Self-standing gender assessments. 4Poverty assessments. 4

Adequacy of Gender Integration into Country Assistance Strategies. 5Consistency with National Action Plans ........................................................................ 6

3. Results Of Bank Assistance For Women . 7

Increased Human Capital . 8Efficacy of Bank Assistance in Increasing Human Capital of Women ........................ . 8Sustainability of Results in the Health and Education Sectors .................................... 12Institutional Strengthening in the Health and Education Sectors ................................. 13

Increased Participation in Economic Development .13Efficacy of Bank Assistance in Increasing Women's Participation in

Economic Development ............................................................................ 14Increased Access to Economic Opportunities .14Increased Access to Credit Facilities .18Increased Access to Training or Skills Upgrading .19Sustainability of Results in the Economic Participation Dimension .19Institutional Strengthening to Increase Women's Participation in Economic

Development ................................................................... 2............................: 20

Institutional Development Impact .20Strengthened Institutions .............................................................................................. 21Increased Involvement of NGOs/Community Groups in Development .21Increased Availability of Gender-Disaggregated Data ................................................ 22

4. What Worked Well? What Needs To Be Improved? . 23

What Worked Well? . 23Country Ownership .23Strategic support at country level .25Partnerships .25

What Needs To Be Improved? . 25Improved policy regime .25Better integration of gender considerations into Bank projects .26Systematic monitoring of results .26Needfor greater support to Bank staff .26

5. Conclusions And Recommendations .27Overall conclusions .27

References ............................... . ...... ......................................................................... 31

Contents (cont'd)

Boxes

1. Relative Position of Evaluation Countries in terms of GDP and UN GenderDevelopment Index (GDI) ............................................................. 2

2. Gender Awareness in Ten Recent Sector Strategy Papers (SSPs) ................................... 43. Has There Been Perceivable Progress After OED Evaluation 2000? ............. ................ 74. Did Quality of Education Improve as Gender Disparity Reduced? ................................. 95. Sustainable Achievements in the Education Sector in Zambia ................. .................... 106. Unanticipated, Unrecognized, and Undocumented Environmental Impact of Men

and Women ............................................................ 117. Do Microfinance Projects Improve Gender Relations? ........................... I ..................... 198. Gender Disaggregated Data ............................................................ 229. Characteristics of Good Practices in Gender ............................................................ 2310. Implementing Agencies Can Save the Situation ........................................................... 2511. Messages from Regional Gender Workshops ............................................................ 27

Tables

1. Did the Bank Do the Right Things in Gender? ............................................................. 32. How Did Bank Assistance Increase Women's Human Capital? ..................................... 83. How Did Bank Assistance Increase the Economic Participation of Women? .............. 144. Integration and Monitoring of Gender Issues during Adjustment .............. .................. 165. How Did Bank Influence Institutional Development for Gender? .............. ................. 216. Implementing Arrangements are Critical in Successful Results in Gender .................. 247. How Did Bank Assistance Impact Country Level Development Effectiveness? ......... 28

Annexes

I. Design and Methodology for Gender Evaluation ................................................... 37II. Analysis of Ten Selected Bank Sector Strategy Papers (FY 1997-200 1) ................ 39

III. Allocation of Resources Across Sectors in Twelve Countries ................................ 42IV. Projects Reviewed for the Evaluation ............................................................ 44V. Socioeconomic Indicators Used For The Evaluation .............................................. 49

VI. National WID Policies in Twelve Countries ........................................................... 53VII. Results of Staff Survey ............................................................ 56

VIII. Project Appraisal Documents Review ............................................................ 59IX. Women's Eyes on the World Bank Evaluation of Human Development Projects in

Latin America ............................................................ 61X. Gender and Adjustment in Ten Countries ............................................................ 63

XI. Report from Committee of Development Effectiveness (CODE) ............ ............. 67XII. OED Presentation at the CODE Meeting ............................................................ 70

XIII. Executive Summary of the Gender Sector Strategy Paper ................. .................... 73

Attachments

1. Recommendations of OED Evaluation 2000 .772. Management Response to the Recommendations of the OED Gender Reviews 79

i

TERMS AND DEFINITIONS

Approach. The conceptual tools the Bank adopts to achieve its gender-related goals. In thisreport, approach refers not to implementation strategy but to broader strategic concepts, suchas a Women in Development (WID) approach or a gender approach, that the Bank uses toachieve its objectives.

Bank Procedure (BP). A statement that specifies procedures and documentation requiredfor Bankwide consistency and quality.

Entry Workshop. A workshop held on September 8-9, 1999, to discuss the approach forOED's gender evaluation.

Fiscal Year (FY). The period from July 1 to June 30 of any year.

Gender. Differences created by social (including cultural, religious, and political) constructsthat result in different roles for, and power relationships between, men and women. Suchroles are learned, vary across different societies, and change over time. However, in thisdocument, it is also used as commonly understood within the Bank, as a surrogate for 'sex.'

Gender Analysis. Analysis that focuses on understanding and documenting the differencesin gender roles, activities, needs, and opportunities in a given context. Gender analysisinvolves the disaggregation of quantitative data by gender as well as qualitative analysis oflaws and institutions. It highlights the different roles and learned behavior of men andwomen based on gender attributes, because these vary across cultures, class, ethnicity,income, education, and time; thus, gender analysis does not treat women as a homogeneousgroup or gender attributes as immutable.

Gender Aware. A term describing interventions or activities that have considered gender orWID issues.

Gender and Development (GAD). An approach that bases the design of interventions on ananalysis informed by gender differences and that targets fundamental social, economic,political, legal, and cultural factors that restrain and or prohibits women from equitableaccess to the benefits of development (Kabeer 1994, Jackson and Pearson 1998)

Gender Blind. A term describing an intervention that does not identify or recognize itsdifferential gender-disaggregated impact, if any.

Gender Empowerment Measure (GEM). See Gender Related Development Index.

Gender Mainstreaming. For the purposes of this report, a process in which gender concernsand women's needs and perspectives are fully taken into account in all aspects of Bankoperations.

ii

Gender-related Development Index (GDI). One of the two new indicators of well-beingintroduced in 1995 by the United Nations Development Program (UNDP), a measure thatincludes gender inequality in its overall assessment of aggregate well-being in a country.

Good Practice (GP). A statement that contains advice and guidance on policyimplementation; for example, the history of an issue, the sectoral context, and the analyticalframework may constitute good practice examples.

Operational Directive (OD). A Bank Directive that contains a mixture of policies,procedures, and guidance, gradually being replaced by Operational Policy, Bank Procedure,and Good Practice.

Operational Manual Statement 2.20 (OMS 2.20). A Bank Policy statement pendingconversion of the requirements relating to social assessments that constitutes Bank policy.Since it was not practice for the Board to consider or approve OMSs, the status of theserequirements are clearly short of requirements that the Board has approved.

Operational Policy (OP). A short, focused statement that follows from the Bank'sagreement and broader policies approved by the Board.

Policy. A set of statements that provide specific institutional directions, guidelines, orprocedures across a particular sector or thematic area.

Regional Gender Workshops. Regional Gender Workshops were organized in four regions(Manila (December 1999), Nairobi (May 2000), Quito (October 2000), and Warsaw (January2001)) to listen to the client's perspectives.

Safeguard Policy. A policy that aims to protect the rights of third parties that may beaffected by Bank-financed interventions.

Strategy. Guidelines for institutionalizing a policy and the processes for implementingpolicy (used generally to cover policy, approach, and strategy).

WID Ratings: Starting in 1988, a systematic Women in Development (WID) rating systemwas instituted for all projects. The system rates appraisal documents and the ratings processis carried out and maintained by the Poverty Reduction and Economic Management (PREM)Unit of the World Bank. Three possible ratings exist under this system:

0 Projects with no attention to women/gender;1 Projects with some discussion of gender or WID issues but no specific action toaddress them;2 Projects with concrete, specific activities addressing gender or WID-related issues.

Women in Development (WID). An approach that focuses exclusively on women, theirroles, responsibilities and needs, without regard to the power relationships between men andwomen.

1

"We are aware that we will never be able to report with absolute accuracy on the achievements of ourwork in relation to women and gender. All we can provide is our best judgment, and continually learnabout what seems to have benefited women and'girls, and what has not, or produced negative effects.Nevertheless the discipline of having to report this may force us to be more rigorous in our efforts to

seek out projects which will achieve more positive change for women and girls, of direct relevance totheir everyday lives."

"Gender Works " OXFAM Experience in Policy and Practice (1999)

1. The Evaluation Context

1.1 The Bank's 1994 Operational Policy (OP 4.20) on "Gender Dimensions ofDevelopment" aims to "reduce gender disparity and increase the participation of women inthe economic development of their countries." It is part of the Bank's poverty reductionmandate.' The policy recognizes that for effective and sustainable poverty reduction, thedesign of development assistance needs to address the differential impact of developmentinterventions on women and men. OP 4.20 focuses on integrating gender considerations intooverall Bank assistance to enhance women's participation in economic development. 2 Itmarks the shift from a Women in Development (WID) to a Gender and Development (GAD)approach within the Bank. This resulted in a change of focus-from one based on separateWID components to one that integrates gender considerations more broadly into Bankassistance-to enhance women's participation in economic development.

1.2 The questions for the Operations Evaluation Department (OED) gender evaluation are:

* Has the Bank's gender strategy been relevant at the country and global levels?* Has the Bank effectively integrated gender into its country assistance strategies and

programs?* How effective has Bank assistance been in terms of sustainable results, impacts, and

institutional strengthening in the area of gender?

1.3 A previous OED evaluation 2000 (Integrating Gender into World BankAssistance)answered the first two questions in the context of a large sample of countries. It concludedthat the Bank's gender policy needed to be clarified, that the responsibilities andimplementation arrangements needed to be strengthened, and that systematic monitoring ofpolicy implementation and progress on the ground needed to be established.Recommendations emerging from this previous evaluation are presented in Attachment 1.The current evaluation chooses the country as the unit of account since it has become thefocus for the Bank's strategies as well as for managerial accountability. This report addresses

'The Summary of Board Discussions (July 1994) on 'Enhancing Women's Participation in EconomicDevelopment' indicates that the Chairman said that the gender focus would support the Bank's poverty reductionobjective and would not be seen as a separate initiative.2 The OP is based on the Bank's Policy Paper of 1994-Enhancing Women's Participation in EconomicDevelopment (referred to as the Bank's 1994 Strategy Paper). Seep. 4, section stating that the strategy aims toenhance the participation of women in development.

2

all three questions in the context of twelve selected countries3 , and evaluates the results of thegender strategy between FY 1990 and FY 2000 (see Box 1).4 In these countries, theevaluation examines all Country Assistance Strategies (CASs) and economic and sector work(ESW) prepared during the evaluation period, and all 180 investment and adjustment projectsthat closed between FY 1995 and FY 2000, and which had hnplementation CompletionReports (ICRs).5 In answering the third question above, the evaluation takes into accountthree dimensions:6

* Increased Human Capital: To what extent did the Bank help to reduce gender disparitiesin the health and education sectors?

* Enhanced Economic Participation of Women: To what extent did the Bank increase theparticipation of women in economic activities?

* Strengthened Institutions: To what extent did the Bank influence institutional changes thatsupport the advancement of women?

Box 1: Relative Position of Evaluation Countries in terms of GDP and UN Gender DevelopmentIndex (GDI)

The evaluation countries (indicated below) were well-distributed relative to other low and middle incomecountries of Bank operations (for which data were available) in terms of gender disparities as indicated byGDI values and per capita income (PPP-adjusted, 1998)(see graph). Most of the evaluation countries liein the low or low-middle income range, except for Poland which has the highest per capita income.

Per Capita GDP and GDI for Selected Countries

085 Poland xxx

The Phlie ( x

O 065 V x ieriam* x> x x

CD 045 - xZis xx %Zan9adq XO¶Coted'lvoire

the GaTrbiaxx

0.25 -6 7 8 9 1V

*Evaluation Countries | Per Capita GDP (bg values)x Otherl

Note: Kyrgyz Republic does not have a GDI value, but, for purposes of graphical representation, is given a value equal to theaverage GDI of the two countries, which are closest in terms of the raw data on different components of the GDI.

'Bangladesh, C5te d'lvoire, Ecuador, Gambia, Haiti, Kyrgyz Republic, Philippines, Poland, Sri Lanka, Vietnam,Yemen and Zambia.4 The evaluation design is summarized in Annex I. The 12 countries have been selected based on regionalrepresentation and levels of gender disparity. These countries are also representative of Bank assistance andinclude countries with little, moderate, and high levels of gender integration into their country portfolios.Another reason for the selection has been a demand for the evaluation from the respective country team..5 In Kyrgyz and Vietnam, with relatively new portfolios, the evaluation has also included ongoing projects.

3

1.4 This report synthesizes the key evaluation findings in these twelve countries andpresents lessons and recommendations. Chapter 2 answers the first evaluation question on therelevance of Bank assistance in these 12 countries. Chapter 3 answers the second question onintegration of gender considerations into Bank assistance, focusing on the results of suchassistance in the three dimensions stated above. Chapter 4 presents examples of goodpractices that emerge from this evaluation. Chapter 5 presents the conclusions and lessons ofthe evaluation. Annexes provide information on methodology and other backgroundinformation. Attachment 1, the recommendations of the previous OED Report 2000.Attachment 2 presents the Management Response to OED recommendations in the previousreport as well as the recommendations presented in this report.

1.5 Regional Gender Workshops were organized in four regions (Manila (December 1999),Nairobi (May 2000), Quito (October 2000), and Warsaw (January 2001)). The goal was tolisten to clients and to understand their perspectives on Bank assistance. In addition,beneficiary assessments were undertaken for this evaluation in Ecuador, the Philippines, SriLanka, Vietnam, and Zambia. Brief missions and field visits were conducted in Bangladeshand Poland. The detailed proceedings and findings from beneficiary assessments areincorporated into this overall report, and will be published separately.

2. Relevance of Bank Assistance

2.1 This chapter builds on the findings of the previous OED report and assesses therelevance of the Bank's gender strategies in the 12 countries. In assessing relevance, theevaluation examines whether the gender dimensions of Bank assistance were: (i) based onsound gender analysis; (ii) satisfactorily integrated into the latest CASs; and (iii) consistentwith country level WID/gender policies. Table 1 presents the evaluation ratings on relevance.In addition, the evaluation examines ten sector strategies issued between FY 1997 and FY2000 to determine the extent of institutional gender awareness (see Box 2).

Table 1: Did the Bank Do the Right Things in Gender?

Criteria D ' o *

N

Soundness of Underlying H SU M H N N N SU N M M SUAnalytical WorkAdequacy of Integration of SU M M SU M N N SU SU SU M SUGender Issues in Latest CASConsistency of Bank Gender H M N H N N N M M H H MStrategy with Country PoliciesOverall Rating for Relevance H M M H N N N SU M SU M SU

H/High; SU/Substantial; M/Modest; N/Negligible

6 These three dimensions are based on the Bank's 1994 gender strategy paper (the basis for OP 4.20) and theWDR of 1990.

4

Soundness of Analytical Work

2.2 The evaluation examines all ESW between FY 1990-2000 to ascertain whether theBank had diagnosed women/gender issues in the twelve sample countries focusing on thecomprehensiveness of the analysis and the adequacy of recommendations.

Box 2: Gender Awareness in Ten Recent Sector Strategy Papers (SSPs)

Ten recent SSPs were examined in order to understand the Bank's analytical work at a global level.The evaluation found that the Bank is gender-aware in sectors related to education, population, andsocial protection. In the area of rural development, the gender awareness of a previous 1997 reportwas lost in the 1999 update. In the other 6 strategy papers, gender issues were virtually absent. ThePublic Sector (Governance) strategy states that issues related to gender, military expenditure, andhuman rights are discussed in other papers, so as not to lose focus. The private sector (PS) and smalland medium sized enterprises (SME) strategies do not discuss gender issues. The former sees thepoor as a homogeneous group, although it mentions the SMEs as particularly important for creatingjobs for the poor (para .3.9 and Annex 1, p. 21). The latter argues that SME workers and owners areunlikely to be the poorest, and that therefore these institutions are not seen as instruments ofpoverty reduction (p. 8). Overall, the findings of this examination of the ten global strategiesreiterate the need for serious introspection on the scope, degree of integration, and implementationof the Bank's Gender strategy within the institution (see Annex II for list).

2.3 Integrating gender into analytical work. Integration of gender analysis into the fullrange of sectoral analytical work has been weak. Most education and health sector workexamines issues related to low enrollment of girls, and maternal and family health. However,they are weak in considering gender issues in other areas. In the rural development andagricultural sector work, consideration of women's/gender issues is limited and tends toignore differential access to and control over assets and resources. Where it does, the analysisdoes not result in appropriate recommendations; two exceptions are in Yemen and Zambia.The widest gap in knowledge is related to women's participation in economic development.Finally, the evaluation finds no consideration of women's/gender issues in ESW related toprivate sector development, small or medium sized enterprises, or public sector management.

2.4 Self-standing gender assessments. The Bank undertook special WID assessments infour countries in the early 1990s (Bangladesh, Gambia, Cote d'Ivoire, Zambia), and inanother two in the late 1990s (Ecuador and Yemen). In five of these six countries, gender wasa critical issue (see Box 1). Overall, these WID/gender assessments are highly satisfactory,i.e. they are comprehensive and cover issues related to all three dimensions listed in para. 1.3.Their analytical frameworks for the most part adopt a gender analysis. They provide practicalrecommendations and suggest gender disaggregated monitoring indicators.

2.5 Poverty assessments. Poverty assessments (PAs) have addressed gender issuessatisfactorily in Bangladesh, Cote d'Ivoire, Gambia, Poland, Vietnam and Zambia. In Haiti,the PA includes an excellent annex on gender, but does not merge the findings into the mainassessment; in Yemen, the gender analysis is deficient. PAs in Kyrgyz (1995), Poland (1995),and Vietnam (2000) also contained discussions of gender and poverty issues, and the first two

5

concluded that gender was not a significant variable in determining poverty. In Poland andVietnam, these analyses provided in-depth and comprehensive coverage of relevant genderand poverty issues, based on information obtained from Living Standards MeasurementSurveys (LSMS). Analysis in Kyrgyz, also based on an LSMS survey, was inadequate. In thePhilippines, Sri Lanka, and Ecuador, the PAs did not discuss gender and poverty issuessatisfactorily, mainly focusing on female-headed households.

2.6 Other weaknesses include timeliness of diagnosis and dissemination of findings.Analytical work is not always undertaken in a timely manner or updated. In Ecuador, Yemen,and Vietnam, satisfactory gender analysis was undertaken only in the late 1 990s, reducing therelevance of the analysis for the decade under consideration. Findings are also notdisseminated well. The Gambia, C6te d'Ivoire, and Zambia WID assessments remained asinternal documents. In Poland, even women's non-governmental organizations (NGOs) wereunaware of the in-depth examination of gender issues in the Poverty Assessment. Recentlythe Bank adopted a more participatory process in preparing women's/gender assessments.Client stakeholders, including women, are more visible in the preparation of recent ESWs. InYemen, the Government-with the Bank and other donors providing support-led thepreparation of the 1998 WID Assessment. In Ecuador, the 1999 gender assessment wasprepared with substantial stakeholder involvement. The process of preparing these reportsresulted in increased gender awareness and capacity within the country, enhanced therelevance of the reports for the client, and ensured timely dissemination.

2.7 In general, there is considerable scope for improving the comprehensiveness andtimeliness of diagnosis, for taking the analysis out of traditional sectors, and for bettermainstreaming of gender issues into the Bank's analytical work.

Adequacy of Gender Integration into Country Assistance Strategies (CASs)

2.8 Gender strategies in the most recent CASs (as of FY 2000) were evaluated according tothe following criteria: (i) comprehensive understanding of critical gender issues; (ii)measurement of implementation progress as reflected by indicators and benchmarks; and (iii)realism of the gender strategy proposed, given the country context.

2.9 Overall, this evaluation confirms OED's previous findings that the treatment of genderin CASs has improved significantly over the evaluation period. In countries with greatergender disparity in educational enrollments, the latest CASs cover a wider range of issues andbetter reflect the critical nature of gender issues. This is true in Bangladesh, C6te d'Ivoire,Gambia, Haiti, Yemen, and Zambia. Outside the social sectors, however, genderconsiderations have not been as well integrated into the overall analysis, and monitoringindicators are not evident even in the latest cases.

2.10 In the other countries, there is greater discussion of gender issues as compared withprevious CASs. The Poland CAS argues that gender is not a variable for determining poverty,and identifies two issues-suggesting that these are likely to be addressed at the project level,but leaving unclear just how this would be done. The latest Vietnam CAS supports theimplementation of the country national action plan for women, which the evaluation sees as afavorable trend. The Kyrgyz 1998 CAS argues that gender is not a significant issue in thecountry. This is surprising since the Asian Development Bank had published a self-standingassessment the previous year, identifying many gender issues based to some extent on Bank-

6

financed household survey results. The latest Sri Lanka CAS also proposes actions foraddressing labor market and gender issues, arguing that gender is not a significant issue in thesocial sectors, except in specific regions. Although the 1996 Ecuador CAS does not considergender issues in a meaningful manner, the latest Progress Report (FY 2000) indicates a morecomprehensive approach to addressing gender issues. The earlier Philippines CAS reflectedconcern with gender issues, but these concerns disappear in the later 1999 CAS. Overall,while there is significant scope for strengthening the quality and consistency of the analysis,there is no doubt that increasingly gender considerations are being integrated into CASs.

Consistency with National Action Plans (NAPs)

2.11 Are the above strategies as reflected in the latest CASs consistent with country levelpolicies or action plans? All twelve sample countries had gender policies or action plans.7 InBangladesh, CMte d'Ivoire, Gambia, Haiti, Philippines, and Sri Lanka, WID/Gender policiesor action plans have existed from the late 1980s. In the other countries, they were more recentand tied in with the Beijing Declaration and Platform for Action. Overall, the evaluation triedto compare whether the Bank's gender/WID objectives were consistent with country nationalaction plans. The evaluation found high degrees of consistency in four countries (Bangladesh,Gambia, Vietnam, and Yemen). In Bangladesh and Gambia, the Bank's assistance wasdirectly responsive to the country national action plans. In Vietnam, the Bank strategy aimedto support the implementation of the national action plan for women. In Yemen, Bankassistance addressed issues of relevance to the country in health, education and agriculture. Infour countries consistency was modest (Cote d'Ivoire, Poland, Sri Lanka, and Zambia). Bankassistance, for the most part, ignored existing gender/WID action plans and integrated genderconsiderations into Bank assistance in an ad hoc manner-and not always doing so in sectorsof priority for women. In four other countries (Ecuador, Haiti, Kyrgyz Republic, Philippines),the Bank did not address gender issues in many relevant areas of Bank assistance that wouldhave been within the scope of the country's action plans and priorities.

2.12 Overall, the relevance of the Bank's gender analysis is modest. It is highly orsubstantially relevant in four countries, modest in five, and negligible in three. The last twosets mainly reflect the loss of missed opportunities rather than clear inconsistencies withcountry policies or action plans.

7 The Beijing Declaration and Platform for Action (para. 297) calls for such national gender/women's actionplans. At least 95 Bank client countries have submitted national action plans to the UN as of April 2000.

7

3. Results of Bank Assistance for Women

3.1 The assessment of results was primarily based on a desk review. It examined allinvestment and adjustment projects with ICRs that closed between FY 1995 and FY 2000 (seeAnnex IV for list of 180 projects assessed for the evaluation). Ongoing projects wereexcluded because evaluating results through a desk review would have been impossiblewithout ICRs.8

3.2 What did OED measure? The gender policy was adopted in 1994. The relevantprovisions in the 1984 Operational Manual Statement (OMS)9 and the Operational Directive(OD) on poverty (1991) focus on women rather than gender. Projects reviewed were preparedbefore 1994. The evaluation, therefore, measured results as they influenced situations forgirls/women and did not focus on changed gender relationships. In addition, the evaluationexamined fifty-three Project Appraisal Documents (PADs) recently approved by the Board toassess whether the Bank had shifted from a WID to a gender analytical framework (see Box 3for findings and Annex VIII).

Box 3: Has There Been Perceivable Progress After OED Evaluation 2000?

* The previous evaluation found that 70 percent of the CASs after FY 1997 integrated genderconcerns better than those produced earlier. This evaluation examined the 57 latest PADs in thetwelve countries and found that this positive trend in the CASs was increasingly reflected at theproject level.* PADs including activities for gender/WID have increased compared to FY 1999. About 45percent included WID/gender activities; and another 10 percent had analysis without an activity.One-third of those had gender analysis, and two-thirds had a WID analysis.* Eighty-nine percent of the PADs indicate some participatory process in project design, 35percent of which explicitly refer to the participation of women; 65 percent of the PADs indicateNGO participation.* Gender-disaggregated monitoring indicators are present in 31 percent of the PADs, while oneproject proposed to measure gender impact.v In less than 10 percent of the projects, the Bank missed a good opportunity to integrategender/WID considerations; 16 percent of projects with male TTLs had WID activities, while 14percent of projects with female Task Team Leaders QITLs) had WID activities.

3.3 Forty-eight percent of projects reviewed had a PREM WID rating system of '1' or '2'(see Terms and Definitions). In 44 percent of these projects, ICRs said that gender was 'notapplicable' as an objective. Only 20 percent of the ICRs incorporated a satisfactory or highlysatisfactory analysis of the results for women. Thus, the desk review did not provide a fullyadequate understanding of Bank results on the ground and were supplemented with selectedfield assessments. The evaluation assessed results in all three dimensions against standardOED ratings of efficacy, sustainability, and institutional development.

8 Supra note 5.9 All projects were prepared after the 1984 OMS 2.20, which recommended that Bank projects-where womenare particularly important participants-should ensure that any adverse impact is mitigated and that results aremonitored.

8

INCREASED HUMAN CAPITAL

3.4 There were education projects in all countries except Kyrgyz and Poland, and healthinterventions in all except Ecuador and Poland. To assess the impact on human capital, theevaluation examined how Bank assistance supported reduction of gender disparity ineducational enrollments and/or improved the health of women.'° Overall, the evaluation findsthe Bank achieved satisfactory results in this dimension. The findings are summarized inTable 2 and described below.

Table 2: How Did Bank Assistance Increase Women's Human Capital?

-e I0 N A3

in~ 0

c ( '09 4

pq ;b) 0 P. P~ >~ N

Outcomes in Education HS U MS S U NA MS NA S HS MU MS

Results in Health HS U NA S MU S MS NA MS HS MS MS

Sustainability in Education HL UL L L* UL NA L NA L L L* L*

Sustainability in Health HL UL NA L* UL NA UL NA L L L* L*

ID in Education H N M SU N NA M NA M SU M SU

ID in Health H N NA M N M M NA N SU M N

Overall Efficacy H N M SU N M M NA M SU N M

*In the present rating system, "uncertain" is not an option; these countries are given "likely" ratings,but they are more on the side of "uncertain."

HS/Highly Satisfactory, S/Satisfactory, MS/Moderately Satisfactory, MU/Moderately.Unsatisfactory, U/Unsatisfactory; HU/ Highly Unsatisfactory.*** H/High; SU/Substantial; M/Modest; N/Negligible.

Efficacy of Bank Assistance in Increasing Human Capital of Women

3.5 Education. In four out of the six countries, with large gender disparities in schoolenrollments (Bangladesh, Gambia, Yemen, and Zambia), gender considerations have beenintegrated into the education projects.1" Bank assistance supported a wide range of demandand supply-based interventions to increase girls' access to education. However, results andquality improvements also varied in all four countries (see Box 4). In Bangladesh andGambia, Bank assistance largely achieved its objectives to increase enrollment of girls and toreduce gender disparity. In Bangladesh, the Bank supported this increase through a numberof means, including inter alia: (i) construction of 30 percent of the estimated requirement ofschools for the decade; (ii) support for improved water and sanitation facilities in schools; (iii)training for more female teachers; and (iv) support for a very successful scholarship programfor reducing the opportunity costs for girls. In areas where schools have been built with Bankassistance, total enrollnent rates for girls increased from 3.53 million (45 percent) in 1990-91

10 The evaluation assessed results in terms of gender disparity to the extent information was available."Haiti and C6te d'Ivoire are the exceptions; here the Bank was focused on addressing physical constraints andsectoral management issues, and did not pay adequate attention to identifying and addressing other supply anddemand factors keeping girls out of school.

9

to 4.75 million in 1996 (47 percent). In Gambia, enrollments for girls in primary schoolsincreased from 40 percent in 1989-90 to 44 percent in 1997. The Bank supported this trendby: increasing the number of female enrollments in the Gambia teacher training colleges from10 to 40 percent; implementing a school building program that resulted in the expectedincrease of seats for primary school children; and through the construction of toilets inschools, which encouraged many parents to send their daughters to school (ICR). Bankassistance also supported the reduction of gender stereotyping and biases in textbooks, andintroduced a gender awareness module into the pre-service teacher-training program. In bothcountries, the Bank was one of many donors in education, and was operating in a context ofstrong Government and other stakeholder commitment to the objective of reducing genderdisparity.

Box 4: Did Quality of Education Improve as Gender Disparity Reduced?

Although most beneficiaries acknowledge that the quality of services has improved after Bankintervention, the evidence on improved learning is mixed. In Bangladesh, an OED audit (1998) of theeducation project points out that only 60 percent of students complete primary school at a cycle costof 7.5 years, while their competency is still at the second grade level. This could be caused byinadequate teacher training and continuing low teacher attendance. In Gambia, student-teacher ratiosincreased from 30:1 to 45:1 through a consolidation of schools, double shifts, and multi-gradeteaching. Although this may have increased efficiency, the impact on quality is uncertain. Additionally,the textbook rental scheme was poorly administered and clearly unsustainable. An OED audit reportsthat enhancing student achievement has been limited and that capacity building was less thansatisfactory. In Zambia, although the average number of teachers increased in urban areas, thenumbers decreased in both project and non-project rural areas. If the quality of education is low, theoverall benefits to greater number of girls in schools is not fully exploited (Ravallion and Datt (1999)).

3.6 In Yemen and Zambia, the results of Bank lending in terms of gender disparity are notevident despite considerable efforts to address the issue. In the former, the Bank focusedmostly on increasing female teachers and increasing residential facilities for girls- measureswhich were perhaps not adequate to influence any significant changes in a highly conservativesociety. The high rate of population growth also affected results, as did the Gulf War, thecivil war, and a major flood. 12 In Zambia, Bank assistance enhanced gender awareness at aninstitutional level with some positive results (see Box 5), but there is no evidence of increasedenrollment of girls in project areas. Enrollments also fell nationally. Field/beneficiaryassessments indicate that in Bank-financed project areas, enrollments fell less than in non-project areas (Zambia Mid-term Review, 1999), but the evaluation could not find impact onfemale enrollments adequate to reduce the existing gender disparity. This was in part becausethe Bank financed more schools in urban areas, where gender disparity was not as high(World Bank, 2000g, Clark 1999).

12 In 1999, in partnership with other donors and led by the Government of Yemen, the Bank has supported athorough diagnosis of the situation of women and children. Perhaps this analytical work will form the basis ofmore effective Bank assistance in Yemen.

10

Box 5: Sustainable Achievements in the Education Sector in Zambia

Although the objective of reduced gender disparity in enrollments was not successfully achieved, therewere other positive results. A Gender Awareness module was introduced into the training curricula foreducation sector management and staff resulting in a number of gender-awareness measures:

(i) Such training strengthened the gender awareness of managers, and created a pool of gender trainers inthe Education Ministry in Lusaka and at the provincial levels. Training programs conducted at all levelsled to increased awareness of gender issues even at local levels.(ii) As a result of such training, students report that boys were being asked to clean classrooms and girlswere being asked to act as monitors-indicating that, in the long term, such training has the potential toreduce gender stereotyping and improve gender relationships.(iii) The District Education Officers formulated a policy of gender balance, requiring the appointment of afemale deputy head, if a school is headed by a male. However, implementation of the policy was difficult,given the lack of female teachers willing to work in rural areas.

3.7 In countries with less gender disparity in enrollments, Bank assistance for educationwas gender-blind for the most part."3 Field assessments suggest no significant adverse impacton girls as opposed to boys in Ecuador, Sri Lanka, and Vietnam. In Vietnam, fieldassessments indicate that the largely gender-blind education project generated more benefitsfor girls than for boys. A number of factors account for the difference, including effectivegender equality policy that combined support for the promotion of equitable participation ofgirls in the sector, relatively better gender-aware local institutions, and poverty targeting. InEcuador, beneficiary assessments clearly indicate that both boys and girls benefited fromimproved educational facilities, and the increased enrollments were proportionate for bothboys and girls. In Sri Lanka, beneficiaries felt that the construction of schools in remote areaswould have benefited girls more because they are most adversely impacted when schools arelocated at longer distances. However, the Bank lost opportunities to address gender biases ineducation, redress issues related to gender stereotyping, and focus on gender-based constraintsto accessing labor markets. In addition, training was often designed for management-malefor the most part. A better-designed program could have targeted teachers who were mostlyfemale.

3.8 Health. There were health interventions in Bangladesh, CMte d'Ivoire, Gambia, Haiti,Kyrgyz, Philippines, Sri Lanka, Yemen, Vietnam and Zambia during this period. Theevaluation finds that project results have influenced the overall positive trends for women inBangladesh and Vietnam. However, Bank assistance focused mostly on maternal health andfamily planning aspects. It did not focus on other gender issues in general in the sector and theevaluation noticed unanticipated impacts on health in a few cases (see Box 6).

" The education project in Vietnam included measures to remove gender biases in text books.

I1

Box 6: Unanticipated, Unrecognized, and Undocumented Environmental Impact on Menand Women

Failure to consider environmental impact can have an adverse impact on women's health. Bankassistance in Ecuador supported projects that financed a number of sub-loans that led to theestablishment of banana plantations and flower production farms. Field visits indicate that about 60percent of the labor consists of women, who complained about the adverse impact of the use ofpesticides on their health. They also suggested that the effluents from these small industries werecontamninating water sources that they then collected and used for household consumption. Men,however, did not see this as a priority issue, raising a concern that these issues may not necessarily beaddressed. Interestingly, the project ICR did not see gender as an applicable issue, while theBorrower ICR stated that gender issues had been pardy achieved. Both contributions did notrecognize the environmental issue.

Source: OED Beneficiary Assessment in Ecuador (September 2000)

3.9 In Bangladesh, national level outcomes such as declining fertility rates, under-fivemortality rates, and rising female life expectancy in the 1990s can be attributed partly to theBank assisted health project. This project (US$756 million) was supported by a large numberof donors who together provided consistent and coordinated assistance to a committedGovernment. The project expanded access to family planning and maternal and child healthfor rural beneficiaries through support for establishment of satellite clinics in rural areas. Awide range of Information, Education, and Communication (IEC) programs devotedsubstantial time to responsibilities of both men and women in family health and gender issues;beneficiary assessments indicate that these are influencing behavior. There is greaterinstitutional focus on gender issues, including: (i) construction and staffing of an Institute forMother and Child Health with 50 percent female staff; (ii) establishing of a Gender Issuesoffice headed by a joint secretary, (iii) expanding of women's role and employment in theHealth Ministry; (iv) developing a gender strategy; and (v) recruiting and training 4,500female health assistants (because experience shows more effective delivery of services towomen by female health assistants). The ICR notes that three inter-sectoral populationprograms were generating positive impacts on women in terms of better family planningmeasures. Additionally, the project shifted responsibility to NGOs increasing their efficiency.

3.10 In Vietnam, the intervention is expected to benefit 50 percent of the country's totalpopulation by improving primary health care services. There are already some indicationsthat the project may be facilitating declining fertility rates through the upgrading of facilities,provision of IUDs, essential drugs, and trained commune and family planning workers.Beneficiary assessments indicate that more women than men used Community Health Centerservices (CHCs) during the year before the field assessment. An actively disseminated IECprogram includes materials on small family size, and equal treatment of sons and daughters.Family planning methods and beneficiary assessments indicate that targeted populations arelistening, although it is too early to talk of impact. The majority of trained workers arewomen. However, beneficiaries complain of the low qualification of staff, higher costs, andCHCs not being open all the time.

3.11 The gender related results also seem to be positive in the other countries, although thedocuments do not provide clear evidence of gender-disaggregated impact. In Kyrgyz, the

12

project is ongoing, but is considered satisfactory by the Government and NGOs, and isexpected to generate positive results for women.'4 In Gambia, Bank assistance helped torehabilitate a number of health facilities, provide them with essential services like water andelectricity, and supply them with necessary equipment for providing better services. Itprovided several housing units in remote areas, addressing a severe constraint to theredeployment of health personnel. It helped to increase the training of nurses-mostlyfemale-in Gambia, and helped improve the curriculum. In the Philippines, Bank assistancehelped to improve maternal health and child services. About 450 midwives were appointed toprovide field services, and there was an indication of increased access to maternal services forwomen. Partnerships with NGOs helped increase the efficiency of services in several remoteand under-served communities. However, in both Gambia and the Philippines sustainabilityof these activities beyond the project period was uncertain.

3.12 In Yemen, project ICRs do not provide much information on results for women-however, the provision of services with a focus on rural areas is expected to have benefitedpoor women. In addition, the project made every effort to increase the number of womenpersonnel in the Ministry, and as health assistants through construction of training facilitiescloser to rural communities with residential facilities for women. However, one ICR notesthat although many women were trained, the public sector was unable to absorb them, thoughsome were able to set up their own services in their villages.

3.13 In Zambia, the Bank-supported, social fund projects financed the construction of anumber of health clinics in rural areas, thereby increasing the access of women to healthservices. In many cases, the project also co-financed residential facilities, addressing a verybig constraint to providing services for the rural poor. Communities were involved in thedesign, implementation, and management of these facilities. Limited field assessmentsindicate that women have spearheaded these activities in some cases, and overall serviceshave improved in such areas. Despite these positive results, in all these cases, the efforts werenot as holistic at a national level as in Bangladesh or Vietnam, and the activities are not likelyto have an impact at a national level.

Sustainability"5 of Results in the Health and Education Sectors

3.14 Sustainability of gender results is likely in the health and education sectors where thereare positive results. This is because of high levels of country ownership in, and commitmentto gender at different levels. Bank assistance was more effective in these sectors because itadopted simple objectives (getting girls into school and improving the maternal health ofwomen) that resonated with the client and were fully owned by them. An approach based onbroader gender issues may have been appropriate, but may have been premature during thelate 1980s and early 1990s. However, there are two concerns that could reduce sustainabilityin the long term. First, principles of cost recovery and the imposition of user fees for socialservices still present a challenge for the Bank. The evaluation finds that Bank assistance hasinvolved cost recovery measures in Cote d'Ivoire, Gambia, Haiti, Ecuador, Vietnam, andZambia without adequate protection of the poorest. In some cases, this was at the behest of

'4 However, it will be difficult to evaluate results because there are no gender disaggregated impact indicators,unlike in Vietnam where 9 out of the 10 project indicators were focused on women's health.'5 Sustainability ratings are based on ICR ratings of overall sustainability adjusted for any country specificgender issues identified by the evaluation.

13

the Government, as in Vietnam. On the other hand, there is evidence of enhanced quality ofservices. Also, in some cases, communities are involved in management. Facility staff have agreater say over how the funds are to be spent, and seem more satisfied with the improvedservices. However, beneficiary assessments in Vietnam, Zambia, Gambia, and Ecuadorindicate that women complained about the costs of services. In the background of theeconomic crisis, one mother in Ecuador said: "When we had to buy textbooks, we had tomake our children go without food." But, there is also contradictory information. A ZambiaSocial Fund Impact study (World Bank 2000) shows that expenditures for education sub-project beneficiaries were 9 percent lower than that of average non-social fund beneficiaries.Beneficiaries spent 3 percent less in sub-project areas than others in non-project areas. Thereis need for greater and closer monitoring to understand whether this perception is indeed areality. Second, improving women's health and increasing enrollments of girls-althoughessential to sustaining growth- are not adequate in and of themselves to result in povertyreduction for women. The Bank focused effectively on increasing girls' enrollments, butfailed to focus attention on issues related to employment for girls entering the labor force aftersecondary education. In the long-term, failure to find jobs for secondary school graduatescould dampen enthusiasm of households to send girls to schools beyond the primary level.

Institutional Strengthening in the Health and Education Sectors

3.15 In both sectors, Bank assistance focused on institutional strengthening-althoughwithout an overall strategy. In the education sector in Bangladesh, Gambia, Zambia, andYemen, there is evidence of better management skills in addressing gender issues anddelivering more gender-aware programs. The Bank has had some influence on the integrationof considerations related to female enrollment at the policy levels in Bangladesh, Gambia,Vietnam, and Zambia. The Bank has strengthened human resource aspects and trained largenumbers of women in all these countries, and there is some evidence that these women haveobtained jobs. In Bangladesh, the health ministry absorbed 4500 female health assistants. InYemen, for example, only a small percentage of the trained women were absorbed by thepublic sector. However, there is evidence that women set up some small clinics in availableplaces in the villages and that the services they provided were welcomed by other women.Female teachers have been trained in all countries, although not always in equal numbers.Removal of gender biases in textbooks occurred in Bangladesh, Gambia, Zambia, and even inVietnam. In Vietnam, there has been significant training in these sectors and because womenpredominate in these sectors, the majority of trainees have been women. The Bank hasincreased partnerships in Bangladesh, Gambia, Kyrgyz Republic, Vietnam and Zambia bycollaborating with NGOs and communities in the design and delivery of services.

3.16 To conclude, the Bank's performance was satisfactory overall, project results weresatisfactory, but the impact on development effectiveness for women was modest in eightcountries, and substantial only in three.

INCREASED PARTICIPATION IN ECONOMIC DEVELOPMENT

3.17 Investment in women's human capital is not adequate to improve the overall welfare ofwomen. Ravallion and Datt (1999), using data from 15 states in India, found that growthcontributed to better poverty reduction in states with better distribution of education andhigher literacy rates, but the positive link is negligible at low levels of education. Moreover,even fornal education is not found to have a positive effect in the absence of reforms that lead

14

to an opening of markets. Thus, in addition to investing in women's human capital, supportingthe increased participation of women in economic activities is a critical dimension for povertyreduction and the enhancement of overall welfare of women. The results of Bank assistancein this dimension are summarized in Table 3 and presented below.

Efficacy of Bank Assistance in Increasing Women's Participation in EconomicDevelopment

3.18 Efforts to address the increased participation of poor women in economic developmenthave been made in only about one-quarter of the projects outside the health and educationsectors. These efforts have been without a strategy, limited to small ad hoc WID components.There is also no indication that either the Bank or the Governments have been able toeffectively upscale or replicate any of these efforts.

3.19 Results of Bank assistance in this dimension were rated along three criteria: (i)increased opportunities for income generation (including where relevant, reduced householdburden increasing time for productive activities); (ii) increased access to credit or otherrelevant economic services; and (iii) increased participation in training or skills upgradingactivities (including literacy programs). Table 3 indicates that except for Bangladesh,Gambia, Poland, and Vietnam, the results of Bank assistance in this dimension areunsatisfactory. In Gambia and Bangladesh, Bank assistance addressed this dimension in aholistic manner touching on all the above issues. In Poland and Vietnam, both men andwomen were able to benefit equitably from largely gender blind Bank assistance becauseimplementing agencies stepped in with targeting mechanisms for women or because bothwomen and men were able to access benefits. In other countries, Bank assistance addressedone or two of these indicators in an ad hoc manner resulting in some localized impact. Theoverall results are discussed below at a project level.

Table 3: How Did Bank Assistance Increase the Economic Participation of Women?

0 °N i EX- ~-~ 0 - .

~0 u N .m

;4 u w N r

Results* MS U MU S U U MU S MU S MU MUSustainability4* UL UL UL L UL UL L L UL L UL ULInstitutional Development*** M N N SU N N M M N M M N

Efficacy M N N SU N N N M N M M M

* HS/Highly Satisfactory, S/Satisfactory, MS/Moderately Satisfactory, MU/Moderately Unsatisfactory,U/Unsatisfactory; HU/ EHighly Unsatisfactory ** HL/Highly Likely, L/Likely, UL/Unlikely, HU/Highly Unlikely*** H/High; SU/Substantial; M/Modest; N/Negligible

Increased Access to Economic Opportunities

3.20 Women constitute the backbone of the agricultural labor in many of these countries.Increased access to services, assets, and capital is very important for increasing theirproductivity and incomes. In addition, given the different roles and responsibilities in the

15

sector and the wide gender disparity in access to land and other resources, the impact ofgender-blind development interventions is bound to be different and often adverse vis-a-viswomen. However, scant attention has been paid to these issues in the projects that wereevaluated. Although 42 of the 180 interventions were in the agricultural sector, only 14projects included components or activities to benefit women. The components were ad hocand appended, but resulted in small positive results for women.

3.21 Bangladesh provides an example of these results. Out of seven projects reviewed in theagricultural sector, four projects included WID considerations. One project included a US$2million component for supporting the economic activities of women's groups. It ultimatelydisbursed only US$200,000 and financed about 16 projects, eight for women. The ICR teamvisited one sub-project and found unsatisfactory results, mainly because of an uncontaineddisease that affected shrimp cultures being cultivated by women's groups. The OED auditteam visited another sub-project and found that it was a good practice worth replicating.Results have been mixed and, in any case, marginal.

3.22 Another project component had two objectives. The first objective (retraining 560women program coordinators) was not achieved due to the cancellation of a Governmentprogram of which it was a part. The second objective was successfully achieved. Inpartnership with NGOs, a gender-focused extension program was formulated andimplemented to enhance homestead production. The ICR states that these activities resultedin increased homestead production and better household nutrition, leading to its replication byNGOs. It also resulted in increased direct contact between women clients and the Departmentof Agricultural Extension. This component was, however, financed by the OverseasDevelopment Assistance. Other projects in Bangladesh dealt with irrigation and targetedlandless women for temporary construction work as part of a popular and effective FoodforWork program. The results are not as positive in other countries.

3.23 The evaluation only examined whether the Bank provided strong and gender-awaresafety nets (see Table 4). Accompanying social safety net interventions were weak inprotecting economic risks. Social Fund projects were implemented in Ecuador and Zambia toprovide decentralized mechanisms to assist the poor through community-initiated andmanaged sub-projects. The projects focused on constructing large numbers of smallinfrastructure in remote areas. None of the projects included a gender strategy. The projectsgenerated temporary employment mainly for men, while in many cases, women provided freecommunity labor. Communities were also burdened in that they had to co-finance theseactivities through monetary or other contributions.

3.24 The evaluation found mixed results of adjustment measures (see Annex X). Forexample, Kumar (1994) suggests female disadvantage in response to adjustment measures,such as the agricultural commercialization that led to the adoption of hybrid maize inZambia. 16 Results indicate that female-headed households have a lower adoption rate ofhybrid maize compared to male-headed households, primarily because the former areconstrained by smaller farm sizes. Kumar also shows that adoption of hybrid maize isassociated with reduced decision-making power for women regarding input allocation, loweraccess for women to proceeds from sale of output, and lower contributions from women tohousehold expenditures on food. At the same time, another study found that laid-off workers

16 Agricultural Marketing and Processing Infrastructure Project (FY 1993).

16

in Zambia face the same unemployment rate as the general population (28 percent), despite ahigher skill base. Women are slightly over-represented among laid-off workers, gravitating toself-employment more than men, and thereby less subject to poverty than men (Kikeri 1998).

Table 4: Integration and Monitoring of Gender Issues during Adjustment

Country Severity of Gender Integration of Gender into Monitoring of Gender Issues inDisparities] Social Protection Measures2 Adjustment Measures3

Kyrgyz M N NEcuador M M NZambia H M MBangladesh H M NVietnam M M NYemen H M NPhilippines M N NPoland N M MSriLanka M M NC6te d'lvoire H M NAverage SU M NNote: All ratings are on a four-point scale, where N Negligible, M = Modest, SU Substantial, H = High.

1. This is based on a relative ranking of counbies based on the UN Gender-related Development Index.2. This rating is provided by the evaluation based on an integration of gender issues into the design of social protection measures

that are financed by the Bank. These measures are not only the ones classified by the Bank as 'Social Protection', but alsoinclude social protection components in other Bank-supported projects of the 1 990s.

3. This rating is provided by the evaluation based on monitoring of gender-disaggregated results in adjustment measures supportedby the Bank in the 1990s.

3.25 Water Supply projects were mostly gender-blind even in countries where access topotable water is an issue for women. Only three out of the eleven projects in the sample hadintegrated gender considerations. Sustainability seems to be a major concern in this sector.At a project level, there have been some positive results. In the Philippines, the involvementof women improved utilization and sustainability of physical facilities. Although the ICRprovides very little information on gender impact in Sri Lanka, an OED evaluation indicatesthat many women were able to diversify their incomes due to time-savings. One-third of theparticipants in community leadership roles were women, and the project brought together acore set of women who work at each level developing women's capacity to participate asactive community members. In some settings, the project led to groups of women startingsavings schemes, and women have been increasingly trained and have begun assumingpositions as technical officers. ICRs indicate that two other projects, although gender-blind,seem to have benefited women. The intervention in Haiti had unsatisfactory results andsustainability was unlikely, but it resulted in about 500,000 people from disadvantagedlocalities being supplied water through kiosks and standpipes. Women must have benefited.Another ICR of a gender-blind project in Yemen points out that women have benefited fromspending less time collecting water. In Bangladesh, although the ICR does not address theissue, the choice of technology (shallow tube wells and low lift pumps) may have resulted inwomen not benefiting from the project because of their relative lack of resources to purchasethem.

3.26 Some gender blind projects have had an adverse or inequitable impact on women'seconomic activities. For example, a bus station constructed in Ecuador had new stalls thatwere leased out mainly to men. It reduced the income of women who used to sell informallyto travelers. A promise from the Municipal Corporation to compensate them nevermaterialized. In Zambia, proj ects intended to increase incomes of the poor led to inequitable

17

distribution of employment and cash income with disproportionate increases in women'sworkload. 7 Cash crops such as coffee and maize promoted by these two agricultural projectswere controlled by men. By extending the production area, women derived minimal benefitsfrom their additional work. Similarly, a market built in Zambia solely for women-withoutconsideration for demand-resulted in the market being more of a community center wherewomen could socialize rather than generate additional income. The advantage was that menwere not interested in leasing the stalls.

3.27 A road built in Zanbia without any community participation benefited both men andwomen, but it increased men's access to alcohol due to the increased transportation flow.Women complained that men spent the evening drinking, while women continued to work. Amarket built in Bangladesh had formal bidding procedures that were often cornered by the'local musclemen' in the vicinities. Women are unlikely to have obtained fair access. Insome projects, benefits went to land owning farmers, while the benefits for landless womenwere in the form of paid construction labor.

3.28 In some cases, gender-blind projects improved women's access to economic resources.11A gender-blind project in Zambia increased employment on coffee farrns. Although only oneof the 76 permanent employees is a woman, during the coffee picking season the majority ofcasual labor was women. An unanticipated impact was that rural women opened up a newwage labor market in the coffee sector. It involved seasonal, sometimes part-time work, andprovided income to women. In some cases, it also opened up access to school and healthservices supported by the coffee farms. Rural men were not displaced in this wage labormarket and generally accepted the change in women's position. Positive changes in the socialposition of women were identified in the case of the Mining Technical Assistance project.Field assessments indicate that the project led to retrenchment of men and the loss of theirformal wage employment. The greater share of providing for their families shifted to women.The women who took care of the homes and ran small businesses prior to the loss of maleemployment had to struggle even more to feed their families. Gender roles changed as menwere reportedly spending greater amounts of time in the household and taking on a largershare of domestic work.

3.29 To increase efficiency, the state scaled back on retirement pensions, delayed the normalage of retirement differentially for men and women, and withdrew programs for theunemployed in Kyrgyz. Reducing 'care' services and withdrawal of social assistance affectedwomen differently than men (World Bank 1992). Women were literally left 'holding thebaby,' constraining their ability to participate in the labor markets and in some cases resultingin taking girls out of school. The Bank did not mitigate this situation even in its safety nets.Another project supported registration of land, a major economic asset. In many oblasts, landwas registered in the name of the male head of the household, thereby eliminating any rightswomen had over land under previous customs of land ownership. The gender impact was not

7 Based on field assessments conducted in collaboration with ISNAR.For example, improved rural roads in some countries have increased the availability of public modes of

transportation making it more accessible and cheaper for the poor, and by providing more income opportunitiesfor rural households. Both the direct and indirect benefits are likely to have impacted women.

18

monitored, losing the opportunity to deepen the understanding of gender issues during thetransition.' 9

Increased Access to Credit Facilities

3.30 There were 25 projects with microcredit components in ten countries during this period.Out of these 16 projects did not integrate any gender considerations and only two involved theprovision for flexible forms of collateral in Philippines and Sri Lanka. Ninety percent of thePhilippines project was canceled, although it had a promising design for benefiting women.The ICR of the Sri Lanka project indicates some good results for women, although overallsustainability was rated as unlikely by the ICR.

3.31 Although the other components or projects were gender blind, field assessments inVietnam indicate that both women and men benefited from the smallholder credit and had arepayment record exceeding 95 percent despite the fact that the project was gender blind.This was due to the involvement of women's NGOs that ensured women's access, as well as alocal regulation that required both spouses to sign as applicants. Women benefited equitably,and that despite the drawbacks of the credit programs (small amounts, application procedures)male and female respondents agreed that the low interest credits have helped improve theirlives and production. Also, the assessment finds that "the increasing role of women in familyaccess to credits (as co-signer or principal applicant) combined with their traditional role asmoney-keeper for the family appear to improve their decision-making power in the family".Second, outreach programs for ethnic minorities in the uplands and mountains are beingstrengthened, thus eliminating the constraints of remoteness and mobility that may beespecially limiting for women.

3.32 There is also evidence of unanticipated impact in some microfmance projects (see Box7). These experiences indicate that the Bank needs to carefully examine expected results,taking into account, in a sensitive manner, the range of possible social and cultural responsesthat may be triggered by women's increased access to micro-finance.

3.33 Gender considerations are not fully taken into account in projects that provide loans tosmall and medium sized industries. Although the then-existing OD on Financial SectorOperations (OD 8.30) permitted targeted credit for women as an exception, the projects didnot diagnose or address gender issues in these sectors except in Gambia and the Philippines.

'9 As late as 1998, the Bank insisted that there were no gender issues, although in 1997 the ADB had publishedan assessment identifying gender issues.

19

Box 7: Do Microfinance Projects Improve Gender Relations?

A Bangladesh poverty alleviation project with a microfinance component is helping tosignificantly reduce poverty. The Asian Development Bank (ADB) in consultation with OEDconducted an independent impact evaluation. A large percentage of the microfinance went towomen, but only 20 percent of the female beneficiaries interviewed had access to money earnedand a voice in spending it. Female beneficiaries confirmed, however, that they felt a greater senseof empowermnent, enjoyed participating in group meetings, and perceived some improvement inthe social status in their households. However, the evaluation also found that families withincreased income were able to withdraw their women from having to work, mainly in the informalsectors, and pay better dowries for marriage - both symbols of prestige for the household and forwomen. In Vietnam too, field assessments indicate that although rural credit to womenitmproves their lives, there is no clear evidence that it changes existing gender orders and divisionof labor within the family. In the Philippines, a few female beneficiaries responded that theirability to generate additional income had led to their husbands taking on responsibility for morehousehold chores, although most felt that it had not significantly alter gender relations.

Increased Access to Training or Skills Upgrading

3.34 Sixty-two percent of the projects mainly in the education, health, water supply, andagriculture sectors included skills upgrading or training components. Only one third of theseprojects integrated concerns for women. ICRs did not usually provide any genderdisaggregated data on the numbers of trainees or discuss the impact of such training.However, the evaluation finds that unless the project design takes gender considerations intoaccount, women are unlikely to benefit from these activities. For example, in Vietnam, fieldassessments show that during implementation the Provincial Agriculture and ExtensionCenter made the selection of women a priority in selecting candidates. This reduced thechances of less women being trained-a likely event in this sector. Large numbers of womenwere trained in the health sector in Vietnam. However, here the project director said: "We donot select by gender, we select by posts, and there are more women in these posts". This wastrue in Poland also. In other countries like Sri Lanka, Philippines, and Zambia, fieldassessments indicate that a gender-blind selection process of trainees (without any designfeatures to ensure gender equity in outcome) is likely to result in majority of men beingtrained.

3.35 In many of these countries illiteracy is a significant constraint to economicparticipation. Yet it was seldom addressed in a meaningful way. There were two projectswith small components. In both cases, the size of the component and the failure to follow upon sustainability issues resulted in women not being able to maintain the momentumachieved.

Sustainability of Results in the Economic Participation Dimension

3.36 Sustainability of results are difficult to aggregate. The evaluation finds that in countrieswith greater gender disparity, sustainability of results for women is unlikely. For example, theZambia Coffee Project resulted in men obtaining formal employment in the coffee farms withsocial security and other attendant benefits. Women, on the other hand, got increased

20

opportunities for income in the unregulated informal sectors, where the risks are higher. Inaddition, in these countries, there are legal and institutional constraints that inhibit women'sequitable access to own and control economic resources. For example, the wife would haveno legal rights to any assets derived from project assistance or from a microcredit facility thatare registered in the husband's name. These larger legal and other constraints- beyond thescope of a single intervention-reduce the sustainability of positive results at the micro-levelin some countries. On the other hand, sustainability for women was likely in Vietnam whereland allotted through Bank assistance was registered in the name of both spouses, and allapplications for microcredit were required to be in the names of both spouses. As the fieldassessments indicate, these measures afforded women greater security.

Institutional Strengthening to Increase Women's Participation in EconomicDevelopment

3.37 The Bank did very little to strengthen institutions to increase the participation of womenin economic activities. In Gambia, it helped to strengthen national institutions that providedsocial and economic services to women. Subsequent audits indicate that the institutions havebeen sustainable. In Bangladesh or Vietnam, where Bank assistance supported micro-financefor women, it merely supported already successful mechanisms for providing loans onaffordable terms to women. In addition, there have been ad hoc activities such as increasingfemale extension workers in the Agricultural and Fisheries Ministries in Bangladesh andYemen, or better partnerships between Government agencies and NGOs in Bangladesh andPhilippines. Overall, the lack of a strategy is apparent and the results have also remainedmarginal and ad hoc. To conclude, the Bank did not effectively address issues related to theeconomic participation of women. Efforts were unfocused and results were consequentlyscattered and unsustainable.

INSTITUTIONAL DEVELOPMENT IMPACT

3.38 The 1994 policy requires the Bank to assist countries to strengthen institutions for theincreased participation of women in development. The evaluation assessed institutionaldevelopment impact along three specific criteria: (i) strengthened development institutions todeliver gender- aware programs (the evaluation assessed whether the Bank strengthenedinstitutions or 'machineries' set up for coordinating women's or gender issues in the country);(ii) strengthened NGOs or community groups to participate in gender-aware developmentinterventions; and (iii) increased availability of gender disaggregated data. Overall, theresults are mixed (see Table 5).

21

Table 5: How Did Bank Influence Institutional Development for Gender?Aspects of Institutional Development

.0,-o- CZ ~ '0 -

Xn .V .70b ia ;

Results in Strengthened Institutions N N M H N N N N N SU M NResults in Better Partnerships with NGOs M N N M N N N N SU M M M

/Communities

Results in Generation and Use of Gender M H SU M N H N H N SU N HDisaggregated Data

Overall Rating for Results M M M SU N M N M M SU M MSustainability** L U L HL U L L L U L U U

Overall Rating for Efficacy M N M SU N M N M N SU N N* H/High; SU/Substantial; M/Modest; N/Negligible ** IHL/Highly Likely, L/Iikely, U/Unlikely, HU/HighlyUnlkely

Strengthened Institutions

3.39 All twelve countries have machineries (term used by Beijing Platform), in some casessince the early 1980s, to coordinate or address women's/gender issues. The Bank has focusedon strengthening these institutions only in Gambia, Cote d'Ivoire, Ecuador and Vietnam.Except in C6te d'Ivoire, these activities have been successful and worthy of replication. InGambia, Bank assistance strengthened the women's machinery to assist in increasing thegender awareness of other important ministries, thereby widening and deepening gendermainstrearning. In 1999, the Bank embarked on an ambitious institutional strengthening ofGovernment agencies in Ecuador to mainstream gender into all the Bank portfolio projects.Assistance was provided by a Population and Human Resource Development (PHRD) grant,and involved highly participatory mechanisms in which project implementation staff led thegender mainstreaming process of their projects. This increased the capacity of internalagencies to deliver better gender-aware programs. In Vietnam, the Bank has initiatedcollaboration with agencies responsible for coordinating matters related to gender equality,and is working to support the implementation of their NAP; there is evidence of positivesynergies that have influenced Bank assistance. However, in most other countries, the Bankhas preferred to deal only with sectoral ministries, thereby contributing to the marginalizationof many of these "machineries." This lack of support raises some concern, given that many ofthese agencies resulted from efforts of UN agencies including the Bank.

Increased Involvement of NGOs/Community Groups in Development

3.40 In 1991, an operational directive recommended that the Bank involve NGOs in itsoperations in order to reach the poor more effectively. About one-third of the projectsincluded NGOs in implementation, particularly in the social and health sectors (Bangladesh,Ecuador, Gambia, Haiti, Philippines, Poland, Sri Lanka, and Vietnam). However, it is notclear whether the Bank focused on the capacity building of these NGOs to bring in the voicesof poor women. There has been some training in Sri Lanka and Gambia, but it often seems tobe supply-driven, raising sustainability issues.

22

3.41 Bank assistance has focused on gender issues at the comnmunity level. However, theimpact is marginal in some countries, because it has not been within an overall equitableinstitutional framework for women's participation. Although 36 percent of the projectsinvolved some form of community participation during the implementation stage, less than 40percent of these gave attention to women's participation. The Bank has required theparticipation of women in community-level institutions in some projects. Field assessmentsindicate mixed results. At one level, such participation increased the burden of women, whowere often responsible for the free labor, while the paid labor went mostly to men. Someprojects required membership of women in committees that managed small developmentactivities. But the outcome was that women were responsible for cleaning or collectingmoney, but rarely had a role in decision-making. Recently in Zambia, one project requiredthat women should be involved in the decision-making team, and not just participate asmembers at large. Such mandatory provisions have some positive impact. Field assessmentsin the Philippines, Zambia, and Ecuador indicate that such prolonged exposure provideswomen with opportunities for leadership, although most preferred to stay in the background.At the same time, in containing Bank assistance to the micro-level, the results have been few,with no impact at the national levels.

Increased Availability of Gender-Disaggregated Data

3.42 The 1984 OMS and the 1994 OPspecifically require the Bank to monitor thegender impact of its assistance. It also calls for Box 8: Gender Disaggregated Datathe Bank to provide assistance to countries in In one country, in the same ministry, projectcollecting, analyzing and using gender- managers maintained gender-differentiated datadisaggregated data. The Bank's efforts have for ADB-financed interventions, but failed tobeen marginal in the first case, with negligible do so for the similar WB-financedbenmparginale inD thndefirs cvase,aith negligible interventions. When queried, they respondedimpact (see OED Gender Evaluation 2000 and that while ADB staff insisted on havingAttachment I, see Box 8). The results orientation differentiated data, the WB had never raisedof these projects in terms of close monitoring of tiis issue.gender-disaggregated results is not high.Seventy-four percent of the projects in the social Source: OED GenderMission

and agricultural sectors included monitoring systems of which only about 10 percent includedgender-disaggregated indicators.

3.43 The Bank has supported a number of govermnents in collecting gender-disaggregateddata. In Kyrgyz, Zambia, Vietnam, and Ecuador, the Bank financed this through LSMS. Inthe Africa region, surveys were conducted under the program for Social Dimensions ofAdjustment. CMte d'Ivoire and Gambia were included in the group of seven countries. Asystematic Gender Analysis of the available survey data was prepared and widelydisseminated by the Bank under the aegis of it Africa Region Gender program GAPS. InPoland, the Employment Promotion and Services project introduced a new, sophisticated andgender-aware labor monitoring system into the Central Statistical Agency in Warsaw. Thesurveys made a substantial contribution to improving the Bank's knowledge andunderstanding of relevant gender-related issues, and perhaps in promoting as well thedevelopment of more poor and gender-sensitive national policies (although there is littleexplicit evidence regarding the latter). NGOs and researchers also use such information.

23

However, LSMS surveys are expensive and require significant technical assistance. Inaddition, unless accompanied by qualitative assessments, they provide gender-disaggregateddata, but do not shed light on intra-household gender relations. In most cases, despite theexpense, sample sizes are small and generate national level statistics that are inadequate inproviding an insight into the differences at sub-national levels.

3.44 To conclude, there has been no Bank strategy for institutional development for genderin any country, except in Gambia, CMte d'Lvoire, and Vietnam, but positive results were onlyfound in Gambia and Vietnam. These efforts have been sporadic and, consequently, resultshave been limited to the micro-level with local impact.

4. What Worked Well? What Needs To Be Improved?

4.1 In addition to good diagnosis of the country gender context, the evaluation found thatcountry ownership, a strategic country-level approach, partnerships, systematic monitoring ofresults, and project design were common characteristics of successful projects (see Box 9).

Box 9: Characteristics of Good Practices in Gender

Country Ownership. Where the evaluation uncovers good results for women (Bangladesh,Gambia, Poland, and Vietnam), it finds (i) greater acceptance of gender equality or a very visibleand actively implemented country gender policy or strategy, and (ii) strong commitmnent reflectedby relatively better gender-aware local institutions.

Project Design and Implementation Mechanisms. The Gambia and C6te d'Ivoire WIDprojects had relevant objectives and were designed within a gender analytical framework based onWID assessments. Yet while the Gambia WID project generated successful results for women, theC6te d'Ivoire project did not. The critical difference was the design features and theinplementation arrangements. (See Table 6).

Effective Partnerships. The Bangladesh Education and Health Projects have better results. Bothprojects involve significant partnerships. Partnership is also evident with NGOs, thereby increasingcompetition in the delivery of services in these sectors. The successful Sri Lanka Water Supplyproject also resulted from similar partnerships with NGOs, community groups, and the UNDPWater and Sanitation Program.

Systematic Monitoring of Results. The successful projects for women (Bangladesh Education,Bangladesh Population and Health, Gambia Education Project and Sri Lanka Water Supply) hadbetter than average Project Status Reports (PSRs) that involved closer monitoring ofgender/women's issues.

(Source: Distilled from OED Evaluation 2000 and 2001)

WHAT WORKED WELL?

4.2 Country ownership. The evaluation finds that gender-aware Bank assistance that wasresponsive to or framed within the country's agenda for women/gender proved to be a veryimportant factor for successful results. This is true in Gambia, and resulted in many sectoral

24

ministries embracing the program. This outcome is all the more noteworthy when comparedwith a similar project in CMte d'Ivoire, which was not framed within the existing WID policyin the country (see Table 6). This is true at a sectoral level also. Increasing girls' enrollmentand improving the status of women's health was a strong policy in Bangladesh and Gambia.When Bank assistance responded to these policies in a gender-aware manner, the results werepositive and likely to be sustainable. This strengthens the argument that the Bank cannotachieve good results for women merely by integrating gender considerations into its ownassistance. A positive and supportive environment within the country is essential. Strong,visible, and active gender or women's policies have been a common characteristic of astrongly supportive environment, as have strengthened institutions at sub-national levels. Inaddition, the potential adverse impact of gender-blind projects was averted in countries withstrong gender aware implementing agencies in Vietnam and Poland, resulting in satisfactoryresults for women (see Box 10). Vietnam and Philippines provide interesting comparisonsthat indicate that ownership and commitment is necessary at different levels in the country.Gender-blind projects resulted in much better results in Vietnam than in the Philippinesbecause in Vietnam, the local institutions and community organizations seemed much moregender-aware, while in the Philippines such gender awareness was more at the national levels.

Table 6: Imple enting Arrangements are Critical in Successful Results in GenderAspect Gambia WID Project Cote d'Ivoire WID Project

(Satisfactory Results) (Unsatisfactory Results)Project Project Objectives were designed to precisely meet the Not based on GovernmentObjectives objectives of the Govemment's Policy agenda. Agenda.

Analytical Project design reflected a detailed understanding of gender No situation analysis.Framework and institutional issues in the Gambia as diagnosed through a

gender assessment.Project Based on a Social Dimensions of Adjustment (SDA) survey, Based on an SDA survey, but noPreparation preparation included assessment of women's priorities and preparation or links to other

needs, development and testing of IEC materials, and review projects.of NGO activities. Three of the components were closelyrelated to parallel Bank projects in education, health andagriculture, a unique feature. Declines in the Portfoho andchanges in both the agricultural sector staff and WIDMinistry staff resulted in such efforts being unsustainable.

Implementing Implementation entrusted to the sectoral ministries, coupled WID Ministry was placed inAgency/Period with strong inter-ministerial coordinating mechanisms. charge (staff say there was

Gender mainstreamning was effective because it was part of a political pressure to do so).process that improved the enabling (and learning) Project was complex but theenvironment within their sectoral areas of expertise, using project was for a two-year term.existing mechanisms instead of creating parallel ones. TheWomen's Bureau was responsible only for coordination andmonitoring. Recognition that change is long termn in this area,p.roject was fot five years.

Government Inter-Sectoral Coordinating Mechanism that met regularly. Inter coordination mechanismCommitment Highly visible project because of broad ownership sustained failed because there was little

in the post complefion phase, especialy in the adoption of a ownership and govemment didnadonal Gender Policy (1999) not comnmit to meeting technical

capacity gaps.Monitoring Conducted a baseline survey to establish indicators as input No baseline survey.Mechanisms into the monitoring system in the Women's Bureau.

Partnerships Effective partnerships forged with NGOs and with other No efforts at partnership.donors.

25

4.3 Strategic support at country level. In Gambia, the Bank provided holistic assistance(focusing on all three dimensions) within a country-led agenda. Bank assistance includedlinkages between relevant Bank-supported projects. Project-specific objectives wereformulated, in some instances, after a needs assessment of potential beneficiaries (see Table6). With Bank assistance, the government launched an extremely successful EEC programtargeting the general public and community-run radio stations and video halls. These wereparticularly effective mechanisms for the dissemination of advocacy and substantive contentmaterial on women's issues, providing significant spill-over benefits. The project initiated aprocess of collaboration and partnership between the Women's Bureau, other governmentagencies, NGOs, and other donors that lasted beyond Bank assistance. Overall, this holisticsupport in critical areas enhanced the country's development effectiveness.

4.4 Partnerships have not been a strength in Bank assistance. There is some evidence tosupport the hypothesis that projects with partnerships have generated better results (see Box10). The evaluation independently selected good practice projects which had generatedsubstantial positive results for women, specifically the Bangladesh Health and EducationProjects, Sri Lanka Community Water Supply Project, and the Gambia Women inDevelopment Project. All four projects had significant partnerships. The sample is small, butthe best practices were identified from a total of 180 projects.

Box 10: Implementing Agencies Can Save the Situation

Actions taken by Implementing Agencies in Poland helped to reduce the potential adverse impact onwomen:* Jobs traditionally held by women were eliminated under a Bank-financed project. Although the projectwas gender-blind, Telecommunications POLSKA retrained all the women for other positions within thecompany. In 1992, 49 percent of positions within the organization were held by women, and by 1999 thatnumber had decreased to 45 percent. Some women preferred to take retrenchment packages.* As a result of restructuring, the mining sector, many underground workers (all men) and surfaceworkers (mostly women) lost jobs. Compensation packages for underground workers were higher thanfor surface workers, although the package for the latter was the same for both men and women. Bothmen and women were offered training programs. In helping households select the appropriatecompensation package, miner's wives were sent information on the options and encouraged to participatein the decision-making. They were also provided with information on other support for women underGovernment packages. A study indicates that although the male miners received the compensation, theyused 80 percent of the compensation for productive purposes. Less than 10 percent was spent onconsumer items (include reference).

**NGOs in Poland assert that the above cannot be attributed to the Bank or the Government, buthappened due to concerted pressure from women's NGOs in Poland (Warsaw Workshop).

WHAT NEEDS TO BE IMPROVED?

4.5 Improved policy regime. There could be many reasons for the failure to address issuesrelated to women in a holistic fashion, in particular in the economic participation dimension.The lack of training and resources is one factor. The evaluation finds that the failure to shiftto a gender analytical framework is another constraint. A gender lens would have clarified

26

what was required, i.e. not special components or subsidies for women, but addressinggender-based distortions that prevent half of the population from accessing the benefits ofBank assistance. Another reason is that staff sometimes do not pay adequate attention to thesustainability and equity dimensions of growth. Whatever the reason, in failing to address theeconomic policy and institutional dimensions holistically and effectively, the Bank hasreduced the overall efficacy of its contribution to the investment in women's human capital.2 0

4.6 Better integration of gender considerations into Bankprojects. Evidence is clear thatwhere there is meaningful integration of gender issues in project design-particularlyinvolving the participation of women in the design and implementation-there are greaterchances of mitigating adverse impacts and in ensuring better results for women. Gambia isone strong case in point as stated earlier. In countries like Zambia, the participation ofwomen in sub-project design and implementation under the social fund also increased theiraccess to benefits.

4.7 Systematic monitoring of results. This evaluation examined all Project Status Reports(PSRs) available in the system for the reviewed projects to see how gender issues weremonitored during supervision (85 projects), and concluded that the focus on results washighly unsatisfactory for gender issues. Forty-six percent did not rate gender objectives, 35percent indicated that gender was not applicable, 14 percent said gender objectives weresatisfactory, and 4 percent, highly satisfactory. In last two categories, only five PSRsprovided an explanation of the rating. All four best practices fell in the last category. ICRswere examined to assess how they dealt with gender issues. Out of the 131 investmentprojects that closed in these twelve countries, 48 percent had a WID rating system of '1' or'2'. In 44 percent of these projects, the ICRs said that gender was not applicable. Only about20 percent provided a satisfactory or highly satisfactory WID/gender analysis of the results.

4.8 Needfor greater support to Bank staff. OED conducted a survey to understandperceptions of Bank staff on integrating gender considerations into Bank assistance. About30 percent (399) of those who received the survey responded. Although 72 percent ofrespondents surveyed had heard of OP 4.20 and 53 percent of OMS 2.20, only 42 percent and24 percent, respectively, had read them. Ten percent were not aware of the policy. Threepercent felt it was a feminist cause! Fifty-two percent felt they integrate gender into theirwork regularly, while 20 percent of the respondents said that gender is not relevant for theirwork. Importantly, 52 percent of respondents said they needed additional resources, while 32percent said they also needed training, and better guidelines if they were to integrate genderconsiderations into their work. The survey indicated the need for a more visible gender policywithin the Bank, and for greater support to staff if they are to implement the gender policybetter (see Annex VII for details).

20 Although investment in human capital mnay have a beneficial impact on households, it is not adequate by itselfto empower women or to influence positive changes in gender relationships. In fact, increased human capital isnot found to have a positive effect on income, unless it leads to increased economic opportunities for the poorand at low levels this effect is negligible (Quality of Growth p. 41). In addition, even with relatively higherlevels of human capital, there would be need to reduce distortions that inhibit women's access to labor marketsor to increase their opportunities for self-employment.

27

5. Conclusions and Recommendations

5.1 Overall conclusions. The study focuses on twelve countries, analytical work preparedduring the evaluation period, and 180 projects. While the findings and lessons can only beindicative in light of the small sample, the countries selected provide regional and incomediversity, as well as differences in the degree of gender disparity. The findings are also*consistent with the other independent impact evaluations. For example, Annex IX, provides asummary prepared by the Women 's Eyes on the World Bank of an evaluation undertaken ofthe impact of Bank Assistance in Latin America. Box 11 provides key messages emergingfrom the regional gender workshops.

Box 11: Messages from Regional Gender Workshops

On Approaches ...* "Do not see us as powerless and vulnerable victims. Empower us (women) with resources, human

capital and assets. We will change our gender relationships."( Africa Workshop in May 2000)* "Why are individuals the unit of analysis? The family should be the unit for analysis and the policy

should aim to ensure equitable resource allocations within a strong family unit..." (Former PrimeMinister, The Honorable Hanna Suchocka of Poland at Keynote Address, ECA Workshop in Warsawin January 2001)

On Bank Performance ...* "For more than 25 years, the Bank has been discussing gender, poverty, and agriculture issues. When

will it stop talking and address issues of inequitable access to assets, resources, and services inagriculture? (Africa Workshop Participants)"The Bank has ignored gender issues in East Asian countries because there is no typical genderdisparity in education enrollments. If this is the Bank's entry point for addressing gender issues, it isirrelevant for East Asia and needs to be changed." (East Asian Gender Workshop Participants)

* " The Multilateral agencies, and particularly the World Bank, came late into the area of gender." Thebilateral and UN agencies were here before." (Latin American & Caribbean Gender WorkshopParticipants)

* "When will the Bank recognize the adverse impact of transition on women?" (ECA WorkshopParticipants)

* "We are grateful that the Bank has finally contacted us-the agency for dealing with gender issuessince 1985 in this country." (ECA/LAC/Afr/EA workshops)

5.2 Bank assistance addressed women's/gender issues in the human development sectors,particularly in countries where there is gender disparity in education enrollments. Overall,results were satisfactory. However, Bank assistance did not focus similarly on increasing theeconomic participation of women or in strengthening the institutional framework for gender.The efficacy of Bank assistance has been substantial only in Bangladesh, Gambia, Poland,and Vietnam, modest in Ecuador, Kyrgyz, Philippines, Sri Lanka, and Zambia, and negligiblein the CMte d'Ivoire, Haiti, and Yemen. In Bangladesh and Gambia, the Bank was animportant player in delivering the positive results. In Vietnam, the Bank can take some credit;in Poland, only minimal (Table 7 provides the overall ratings).

28

Table 7: How Did Bank Assistance Impact Country Level Development Effectiveness?

4J~~~~~~~~~~~~~~(

0" 0

Efficacy of Bank Assistance-

Relevance of Bank Assistance* H M M H N N N SU M SU M SU

Results of Bank Assistance** SU N M SU N M M M M SU M M

ID Impact* M N M SU N M N M N SU N N

Overall Efficacy* SU N M SU N M M SU M SU N** M

*H/High; SU/Substantial; M/Modest; N/Negligible;**The lack of sustainability was a major factor in determining the overall efficacy rating for Yemen.

5.3 Although causality is difficult to assert, the evaluation finds the following commonfactors in Bank assistance to Bangladesh and Gambia. First, country ownership andgovernment commitment is evident as reflected in active policies and action plans for women,and by strong and accountable counterparts. Second, the Bank provided assistance consistentwith the country's own action plan, thereby increasing the likelihood of developmenteffectiveness and sustainability. Third, there was convincing analytical work thatunderpinned and laid a strong technical foundation for Bank operations in both countries.Much of this analysis and diagnosis was within a gender analytical framework. Fourth, inthese countries, the Bank adopted a holistic approach to gender issues. It not only focused onintegrating women's/gender considerations into the education and health sectors, but alsoincluded interventions that would increase the participation of women in the economicsectors. Fifth, Bank assistance addressed key institutional aspects through its assistance, suchas building institutional capacity of development agencies to deliver gender-aware programs,and by raising the interventions to a higher level than just tinkering at the margins to help afew women. Sixth, in both countries, there were incentives for strengthening genderawareness at the community levels. Seventh, non-governmental organizations were involvedin implementing Bank assistance. Eighth, in both countries, the Bank is only one actor. Thereis much larger evidence of partnerships in providing development assistance. In none of theother countries is Bank assistance characterized by all the above factors.

5.4 This evaluation reiterates the recommendations of OED Gender Evaluation 2000 thatthe Bank's conceptual and analytical frameworks for addressing gender considerations needurgent clarification. The Bank seems to have focused on addressing gender issues incountries where there is noticeable gender disparity in education enrollments or wherewomen's health status is poor.2' While these are appropriate entry points in Africa, South

2] This is consistent with International Development Indicators, which also measure gender equality in primaryand secondary enrollments as key indicators of progress towards gender equality.

29

Asia, and in some countries in other regions, they did not provide entry points for addressinggender issues in countries with no evident disparity in the social sectors, but with otherimportant gender gaps in economic activities. The Bank needs to clarify how its genderpolicy is linked with its poverty reduction mandate, and explain the operational implicationsfor Bank processes and practices. Clarity of this basic conceptual framework is needed todevelop country strategies and instruments appropriate tools to meet desired policy objectives.

5.5 At the project level, the evaluation finds a tendency for results to be superior for womenwhen gender considerations are integrated into the design and implementation of assistance.22

Such integration into Bank assistance is found to be particularly necessary in countries withgreater gender disparities. In countries with greater participation of girls in education, gender-blindness did not result in constraining access to girls or women in the social sectors.However, even in these countries, the failure to address gender and economic issues resultedin women not benefiting equally or equitably, thereby confirming the need to integrate genderconsiderations into project design. Exceptions were Poland and Vietnam. Better results forwomen resulted from the enhanced gender equality in these countries, and the fact thatgender-aware local agencies intervened in many cases to enhance the impact of theintervention on women. In Vietnam, good poverty targeting also made a difference.

5.6 The evaluation also suggests that the Bank can get the biggest returns for its effortswhere it is able to assist in the creation or transformation of development institutions in theborrowing country that in turn have a pivotal role in delivering gender-aware programs andprojects. The evaluation found that in countries such as Poland and Vietnam with stronggender policies and gender-aware institutions at local levels, gender-blind projects havegenerated positive results for women during implementation. Given that the Bank is movingtowards programmatic lending, strengthening of in-country institutions needs to be the focusof a proactive Bank strategy, if its programmatic assistance is to generate sustainable results.

5.7 OED's Gender Report 2000 recommended that the Bank's gender policy be clarified,the responsibilities and implementation arrangements strengthened, and systematicmonitoring of policy implementation and progress be undertaken on the ground (seeAttachment 1). Based on the findings and lessons of this evaluation, OED furtherrecommends:

(i) Strengthen borrower institutions and policies. Before and since Beijing, mostborrowers have, at the behest of UN Agencies, formulated national policies and/oraction plans for gender equality and/or the advancement of women, consistent with theprinciples of the Beijing Platform for Women. The Bank should strengthendevelopment institutions to support the implementation of these policies/action plans. Incountries where such policies and plans are weak, support for their strengthening(through country dialogue and non-lending services) should be a Bank priority.

22 In addition, the evaluation found that sixty-one percent of all projects with overall results rated 'S' and 'HS'had integrated gender issues; all projects rated as HS had high levels of gender integration. Only 45 percent ofprojects rated unsatisfactory or highly unsatisfactory had integrated gender considerations. The levels of genderintegration were independently validated by the OED evaluation, although it relied on ICR ratings for overallresults.

30

(ii) Integrate gender considerations into country assistance strategies. Based on acomprehensive diagnosis, the CAS should explain how Bank assistance will take intoaccount the linkages between poverty and gender. This assistance and its underlyingstrategy should be explicitly related to the Borrower's policy framework for gender.

(iii) Integrate gender into the design of Bank-supported projects. Gender analysis shouldbe integrated into the economic and social analysis carried out in the preparation anddesign of Bank-supported projects so that both men and women are able to access thebenefits equitably. This is especially critical for countries with high gender disparities.

31

References

World Bank Documents

Blackden, Mark and Bhanu, Chitra. Gender, Growth and Poverty Reduction. 1998 Special Program ofAssistance Status Report on Poverty in Sub-Saharan Africa, World Bank Technical Paper.

External Gender Consultative Group (EGCG). Report of the First Annual Meeting, April 1996.

External Gender Consultative Group (EGCG), Report of the Second Annual Meeting, May 1997.

McAllister, Elizabeth, 1998. "Institutionalising Gender Issues. Effective Change Strategies" WorldBank Unpublished paper), 1998.

Morgan, P. 1998. Mainstreaming Gender in the World Bank: An Organisational Analysis, Preparedfor the Gender Sector Board. Washington, D.C.: World Bank.

Moser, Caroline 0. N., Annika Tornqvust, and Bernice van Broknhorsf. 1998. Mainstreaming Genderand Development in the World Bank: Progress and Recommendations. Washington, D.C.:World Bank.

Murphy, Josette. 1995. Gender Issues in World Bank Lending, Operations Evaluation Department,Washington DC, World Bank.

1997. Mainstreaming Gender in World Bank Lending: An Update. Operations EvaluationDepartment, Washington, D.C., World Bank.

Piccioto, Robert. "The Missing Development Links: Gender and Social Capital". Paper presented atthe Gender and Development Workshop, Washington D.C., April 2, 1998.

Wolfensohn, James D. 1995. Speech at the Beijing Conference.

Dayton, Khan, Ribe, Schneider. 1993. Country Policies for Poverty Reduction: A Review of PovertyAssessments, EDP Discussion Paper Series No. 15, Washington D.C., World Bank.

Fong, Monica; Wakeman, Wendy; and Bhusan, Anjana. 1996c, Toolkit on Gender in Agriculture,Gender Toolkit Series, No. 2, Gender Analysis and Policy, Poverty and Social PolicyDivision, UNDP-World Bank Water and Sanitation Program, Transportation, Water, andUrban Development Department, The World Bank.

Kikeri, Sunita. 1998 Privitization and Labor-Wat Happens to Workers When Governments Divest?World Bank Technical Paper # 16.

Lazreg, Marnia, ed. 1999b, Making the Transition Workfor Women in Europe and Central Asia,World Bank Discussion Paper # 411. Washington, D.C.

World Bank. 1993. Enhancing Women 's Participation in Economic Development. World Bank PolicyPaper. Washington D.C.

32

References (cont'd)

1994. "The Gender Dimensions of the Bank's Assistance" Operational Policy 4.20.Washington D.C.

1995. Toward Gender Equality, The Role of Public Policy. The World Bank.

1995a. Advancing Gender Equality: From Concept to Action. Toward Gender Equality.The Role of Public Policy. Vol. I and II, Washington D.C.

1995b. Gender Issues in World Bank Lending. J. Murphy, Operations EvaluationDepartment, World Bank

1995., Towards Gender Equality: The Role of Public Policy. Development in Practice.Washington DC

1995d. Enhancing the Participation of Women in Development. Washington D.C.

1996, Implementing the World Bank's Gender Policies, Progress Report No. 1.

1996b, Poverty Assessments, A Progress Review. The World Bank

1996d, Toolkit on Gender in Water and Sanitation, Gender Toolkit Series, No. 2, GenderAnalysis and Policy, Poverty and Social Policy Division, UTNDP-World Bank Water andSanitation Program, Transportation, Water, and Urban Development Department, The WorldBank.

1998, Quality of ESW in FY 1998: A QAG Assessment. Quality Assurance Group.Washington D.C.

1998b Country Assistance Strategies Retrospective and Outlook. March 30, 1998,Operations Policy and Strategy. Washington, D.C.

1998c. Environment Matters at the World Bank. Annual Review, fall edition.Environment and Socially Sustainable Network. Washington, D.C.

1998d. Implementing the Ethiopian National Policyfor Women: Institutional andRegulatory Issues. Addis Ababa: The Women's Affairs Office, Federal Democratic Republicof Ethiopia and Washington D.C., World Bank.

1998e. Children and Women in Yemen. A Situation Analysis. Volumes 1-IV, Governmentof Yemen, UNICEF, The World Bank, and Radda Barnen.

1999. Poverty Assessments: A Follow-up Review, Operations Evaluation Department,The World Bank.

1999a, External Gender Consultative Group. Internal Report, Fourth Annual Meeting,November 14-17, Washington, D.C.

33

Reference (cont'd)

1 999c. Gender Analysis and Policy Making for Development. Volume I. World BankDiscussion Paper No. 403. Washington, D.C.

2000. Policy Research Report on Gender and Development. Gender Sector Board andDevelopment Economics. Washington, D.C.

2000a. "Quality of Entry in CY 1998". Quality Assurance Group. Washington, D.C.

2000b. Advancing Gender Equality: World Bank Action Since Beijing. Washington, D.C.

2000c. Gender and Law in Francophone Sub-Saharan Africa. The Role of the World Bank,Findings, No. 155, Economic Management and Social Policy, Africa Region. Washington,D.C.

2000d. "Vietnam Development Report: Attacking Poverty". Country EconomicMemorandum. Washington, D.C.

2000e, Poverty Reduction and the World Bank Progress in Fiscal 1999, Poverty ReductionBoard, PREM

2000f, Engendering Development. Policy Research Report, World Bank,Washington, D. C.

2000g, Social Funds 2000, Impact Evaluation of the Zambia Social Fund, PreliminaryDraft, Chase, R., and L/ Sherbume-Benz

Independent Evaluations of Bank Assistance

Bread for the World Institute. 1996. Gender Justice and the World Bank, Development BankWatchers' Project.

Christen Michelsen Institute. 1999. WID/Gender Units and the Experience of Gender Mainstreamingin Multilateral Organizations: Knights on White Horses. Evaluation Report 1.99. Oslo: RoyalMinistry of Foreign Affairs.

Gender Equity and the World Bank Group: A Post-Beijing Assessment, Women's Eyes on the WorldBank-U.S. in collaboration with Women's Eyes on the World Bank - Latin America & 50years in Enough: U.S. Network for Global Economic Justice (October 1997)

Whitehead, Ann and Lockwood, M., 1999. Gender in the World Bank's Poverty Assessments: SixCase Studies From Sub-Saharan Africa. Discussion Paper. United Nations Research Institutefro Social Development.

34

Reference (cont'd)

Other Evaluations/References

Anker, Richard. 1998. Gender and Jobs: Sex Segregation of Occupations in the World. Geneva,International Labor Office.

Bauer, Armin, David Green, Kathleen Kuehnast. 1997 Women and Gender Relations. The KyrgyzRepublic in Transition. Asian Development Bank, Manila, Philippines.

Buvinic, M, Gwin, C., Bates. 1996. Investing in Women: Progress and Prospects for the World Bank.Washington, D.C. John Hopkins University Press.

Development Assistance Committee. "Workshop on Evaluating Gender Equality and Women'sEmpowerment". November 25-26, Washington, D.C.

Dollar, David, and Roberta Gatti. 1999. "Gender Inequality, Income, and Growth: Are Good TimesGood for Women?" Policy Research Report on Gender and Development, Working PaperSeries No. 1. Washington, D.C., World Bank.

Elson, Diane, ed. 1991. Male Bias in the Development Process. Manchester, U.K.: ManchesterUniversity Press.

Geisler. G, 'Silences speak louder than claims, gender development and agricultural households inSouthern Africa, World Development at 21, 1993. (Cited in Heward below)

Gender Strategy for the Third Zambia Social Investment Fund: Assessment of the Community BasedOrganizations and Participation of Women in MPU Projects (Volume II), First Draft, January18, 2000, submitted by Women for Change & Chalo Environrment and SustainableDevelopment Consultants

Heward, Christine, and Sheila S. Bunwaree. 1999. Gender, Education, and Development: BeyondAccess to Development. London: Zed Books.

Hill, M. Anne, and Elizabeth M. King. 1995. "Women's Education and Economic Well-being."Feminist Economics 1(2): 1-26.

Mike Ingham. 2000. Jobs for the girls? Ending discrimination in the Polish labour market. EuropeanStudies Research Institute University of Salford, October 9, 2000, inhttp://www.id2 .org/static/6ami .htm

Inventory Results of the MDB Survey on Policy Evaluation: A Preliminary Report, September 1999,Office of Oversight and Evaluation, Inter-American Development Bank, September 1999

Jackson, Cecilia and Pearson, Rose. 1998. Feminist Visions of Development, Gender Analysis andPolicy, London, Routledge.

Jahan, Rounaq. 1997. The Elusive Agenda: Mainstreaming Women in Development. London and NewYork, Zed Books.

35

Reference (cont'd)

Kaufmann, Daniel. 1998. "Challenges in the Next Stage of Anti-corruption." In New Perspectives onCombating Corruption. Washington, D.C.: Transparency Intemational and the World Bank.

Klasen, Stephan. 1999. "Does Gender Inequality Reduce Growth and Development? Evidence fromCross-Country Regressions." Background paper to the Policy Research Report on Gender andDevelopment. Washington, D.C., World Bank.

Kumar, Shubh K. 1994. "Adoption of Hybrid Maize in Zambia: Effects on Gender Roles, FoodConsumption, and Nutrition,. " Research Report 100. Intemational Food Policy researchInstitute, Washington, D.C.

Mach, C., Smyth, Ines, and Mukhopadhyay, M., 1999, A Guide to Gender-Analysis Frameworks,Oxford, Oxfam.

Mieke Meurs, Imagined and Imagining Equality in East Central Europe: Gender and EthnicDifferences in the Economci Transformation of Bulgaria, Department of Economics,American University, 1998.

Ostergaard, L, 1992, Gender and Development, A Practical Guide. London and New York, Routledge.

Ravallion, What is neededfor a more pro-poor growth process in India, (Development ResearchGroup, World Bank (an Invited paper for a symposium in honor of Kirit Parikh).

Sen and Grown, 1987, Development, Crises, and Alternative Visions: Third World Women 's'Perspectives. New York. Monthly Review Press New Feminist Library.

Smith, Lisa C., and Lawrence Haddad. 1999. "Explaining Child Malnutrition in DevelopingCountries: A Cross-country Analysis." FCND Discussion Paper No. 60. Washington, D.C.,Intemational Food Policy Research Institute.

Sen, Amartya, 1999. Development as Freedom. New York, Knopf, Inc. Educating Girls: Strategies toIncrease Access, Persistence, and Achievement, Advancing Basic Education and LiteracyResearch Study, USAID, December 1991.

The Challenge ofMainstreaming: A Report to the Board of Executive Directors on the implementationof the WID Action Plan, 1995-97, WID Unit, Sustainable Development Department,September 1998, Inter-American Development Bank.

Thomas, Vinod, Mansoor Dailami, Ashok Dhareshwar, Daniel Kaufmann, Nalin Kishor, Ram6nL6pez, and Yan Wang. 2000. The Quality of Growth. New York: Oxford University Press.

SIDA, 1997, Lessons Leamed from the Primary Education Sector project, Swedish Embassy.

United Nations. 1995. Human Development Report. New York.1999. Human Development report. New York.

Women and Development. Policy and Implementation in Netherlands Development Cooperation (985-1996), Netherlands Development Assistance, Policy and Operations Evaluation Department,Ministry of Foreign Affairs, 1998

36

Reference (cont'd)

The World Bank, 1992, Country Study "Kyrgyzstan. Social Protection in a Reforming Economy".

The World Bank, 1996, Study on Evaluation of Physical Works under the GEP, December, Dhaka,Sthapati Sangshad.

The World Bank, Privatization and labor: what happens to workers when governments divest?1998/02/01 WTP396 World Bank Technical Paper.

37 Annex I

ANNEX I: DESIGN AND METHODOLOGY FOR GENDER EVALUATION

Objective:

To assess gender related results of Bank assistance.

Principal Evaluation Question: /

To what extent did Bank assistanceinfluence gender equity? *Increased participation

of women in economicdevelopment

Evaluation Framework: Improved institutional*Bank Products and Services framework*Results on the ground level

a) Twelve selected countriesb) All analytical workc) All projects that closed

between FY95-FYOO*Bank and Borrower Performance

38 Annex I

IF t _ Stakeholder Assessment in Regional Workshop in

Entry Workshop Philippines and Vietnam East Asia With Gov. ofJuly 1999 Philippines (12/99)

(internal andexternal Stakeholder Assessment in Zambia Regional Workshop in

participants) and Project audit in Gambia Africa with UNDP &

ISNAR (5/00)

_________________ B etteroundtable rStakeholder Assessment in Regional Workshop in Development

Roundtable Ecuador Latin America withwith senior IADB (9/00) OutcomesGovernment__ Goliembernmen Stakeholder Assessment in Poland Regional Workshop in

September 2000 ~~~~~~~~~~~Eastern Europe and Central

I ~~~~~~~~~~~~~~Asia with UNIFEM and; ECE 01/01) ~~~~~~~~~~RevisedStakeholder Assessment in EE0/1 eie

Consultation Bangladesh and lirnited project Genderwith NGOs _ assessments in Sri Lanka and India StrategyOctober 2000 _ Detailed Desk Review

_ ~~~~~~~~~~~~~of Bank Assistance in

l ~~~Co ntinuous Consultation 12 countriesi~ ~ 0 with Gender Sector Board

Desk Reviews (9/99-3/00) Gender Report 2000: Gender Report 2001:27 Country Assistance strategies An Assessment of - An Assessment ofincluding analytical work and lending Gender Integration Gender Results

CONSULTATION PROCESS FOR GENDER EVALUATION

39 Annex It

ANNEX II: ANALYSIS OF TEN SELECTED BANK SECTOR STRATEGY PAPERS

(FY 1997- 2001)

Policy Gender Integration

1999 The strategy has a strong gender dimension although "gender" is very much related to boys and girls only. The level ofEducation gender integration encompasses the policy, guidelines, the working procedures and organizational arrangements.Strategy

1977 Health: The strategy emphasizes the need to "understand more fully the individual and household response to changes" and toNutrition and encourage individuals and households to take more responsibility for their own health (p.6), gender is mentioned oncePopulation and the strategy continues to adopt a traditional approach, with women's issues discussed in relation to education and

fertility. The SSP finds the root causes for failing health policies in the client countries themselves and not in theBank's approach to development. Hence, gender is not recognized as an ignored issue in the Bank's own practice, noris gender an issue that should explicitly be addressed in the health dialogue with other countries. The SSP mentions theCAS as the Bank's central vehicle for development assistance, but nothing is said about gender or women in thiscontext. The regional and country statistics are not sex-segregated except for an indicator for "unwanted fertility rate".Thus policies, working procedures and organizational arrangements at client country level on health do not addressgender.

1998 Population This note states the policy clearly, and articulates instructions and operationality. This strategy has a high genderand the World profile, despite a slight WID bias, in that the role of men in implementing the strategy is not fully recognized.Bank: Adaptingto Change

1997 Rural The strategy clearly acknowledges the lack of attention to gender issues as one critical factor. A gender approach isDevelopment acknowledged essential if the overall aims of the agricultural sector of the Bank are to be met. The major challengesFrom Vision to are reducing rural poverty and hunger, raising economic growth, increasing global food production and halting naturalAction & the resource degradation (pg.3). In the analysis of the question why former Bank efforts have not gained the successes1999 Update hoped for in rural development, gender issues indeed have a place. The SSP repeatedly articulates poor as women and

men, with different needs and with rural women as even more disadvantaged. A lack of expertise and competence in

40 Annex II

Policy Gender Integration

the field of gender is recognized and gender is on the leaming agenda. Nevertheless gender remains optional forimplementation of rural policies at the regional and country level. The gender focus in this SSP is quite high save thatno mechanism for sanctions or rather remedying are proposed. Surprisingly, the 1999 update of the strategy seems tohave taken a step backwards, with no references to gender; and only two references to "women', both times in thecontext of rural people.

2000 Social The draft Strategy was returned for revision by the Bank's Executive Board, which commented inter alia about theProtection inadequacy of attention paid to gender issues. The revised document is significantly considerate of gender, although itStrategy recognizes no differences between rural and urban women and does not address the internal Bank structures, working

procedures and resources. Gender could evaporate in implementation, unless monitored closely.

Small and The Small and Medium Enterprise Strategy (SMES) ignores the economic return on investment on women, and theMedium differential needs of, and impacts of men and women. In part, this can be explained by the strategy's market focus.Enterprise The SMES justifies its absence of any discussion of women by arguing that strategies that try to simultaneouslyStrategy address inequities and market inefficiencies, create a distortion of market forces, result in duplication of efforts and

cover some types of services and clients while missing others. Furthermore, small and medium enterprise (SME)workers and owners, it is argued, are unlikely to be the poorest in the labor force and therefore SMEs are not seen asinstruments of poverty reduction (pg 8). The SMES paper dissents from the Private Sector Development (PSD) paper(discussed below), which mentions the SMES as particularly important for creating jobs for the poor (para. 3.9 andannex 1, p. 21).

2000, The energy sector is not a typical target group, let alone a typical gender centered sector. However, the SSP indicatesEnvironment & (although not in depth) in certain paragraphs the linkages between energy inputs and behavior at macro and nationalEnergy (Fuelfor levels, and the effect on the community level. This makes the document rather interesting and realistic. However, thisThought) line of thinking is not extended a little further by adding the gender dimension. Gender could have been incorporated

particularly in those paragraphs, where the end-users behavior is at stake, and people are in focus, such as on smallsource pollution (p. 27). This would have been not only more consistent with the Bank policy on gender, but thepolicy would have gained in terms of transparency and effectiveness.

2000, Private The PSD strategy (2000) argues that economic growth is essential for poverty reduction and that the private sector isSector the main engine for such growth. The agenda of accelerating development of a market economy, driven by privateDevelopment investment, includes: stimulation of an enabling environment, encouraging private enterprises development and

healthy capital markets, catalyzing extemal capital flows and controlling crises. The absence of any gender analysis isnot very surprising because the SSP considers the poor as a homogeneous group. In this train of thought a gender

41 Annex II

Policy Gender Integration

approach is hard to justify, leaving the strategy unsatisfactory from a gender dimension. For example, developmentliterature documents comprehending the labor market segmentation is essential for Private Sector Development. Thisrequires understanding of gender, as there are typical female or male dominated sectors, which in turn will have itsimplications for policy fornulation. Training suggested in the SSP has gender implications. Creating "an enablingenvironment", which is high on the-PSD agenda, cannot be effectively done if the skewed reality of the sexes is nottaken into account. The Bank lost an opportunity to provide leadership in this area.

(2001) Reforming The policy strategy paper refers only once (in its first chapter pg. 4) to gender equality - along with militaryPublic expenditure, conflict resolution and human rights- as specialized topics of importance to governance. It argues that theInstitutions and SSP does not integrate these issues in order not to lose its focus, and that other studies and reports of the Bank areGovernance treating these topics. Gender is globally recognized as a governance issue (USAID (1993) OECD (1994)), cutting

across all sectors and part and parcel of the political and institutional realities on the ground. The argument in the S SPis a confusing signal that this SSP, compiled under the aegis of PREM, reduces gender to almost a non-issue, whereasthe Gender and Development Group (also within PREM) advocates gender as essential for economic efficiency andpoverty reduction.

(2001) Financial The strategy discusses the poor as a homogenous group; gender issues then becomes hard to justify.Sector

42 Annex III

ANNEX III: ALLOCATION OF RESOURCES ACROSSSECTORS IN TWELVE COUNTRIES

1. In countries with gender disparities, interventions in certain relevant sectors have thepotential to benefit women in particular. The Policy Paper (The Gender Dimensions of theBank's Assistance, World Bank, 1994) suggests some promising approaches to improvingwomen's status and productivity. These include expanding girls' enrollment; improvingwomen's health; increasing women's participation in the formal labor force; expanding optionsin agriculture and management of natural resources; and providing financial services for women.Also, in spite of women's significant role in managing household water and sanitation in manydeveloping countries, a variety of baniers reduce their access to these services. In addition tosupporting these broad themes, in his address to the Fourth UN Conference on Women in Beijing(1995), the President of the World Bank stressed as a priority the need to ensure that women notbe hurt by structural adjustment programs. This implies the importance of lending to broadsectors like Education, Population, Health and Nutrition, Environment, Water Supply andSanitation, Social Protection, Agriculture, and Rural Transport from a gender perspective.'

2. The lending portfolio of the countries in the study reveal interesting patterns (Table 1).First, the share of total Bank lending to the relevant sectors in the 1990s is negatively correlatedwith the country's Gender Development Index.2 This indicates that in countries with highergender disparities, the share of Bank lending to sectors of particular importance to women ishigher. This is potentially beneficial to women. Second, when we compare the change in theshare of these sectors from the 1980s to the 1990s, we find that Bank lending moved in favor ofthese sectors more strongly in countries with higher gender disparities.3 Third, when we examineabsolute lending to the seven relevant sectors combined, we find that this rose in all countriesexcept in Sri Lanka where gender disparities are moderate. Thus, this evidence indicates thatBank lending was moving in the right direction as far as gender was concerned.

3. However, within this overall positive picture, some areas of concern are discernable when welook at the individual countries. Lending to education and health declined in Yemen between thetwo decades-in both absolute terms and as a share of total lending-despite female-maleprimary enrollment ratios of only about 40 percent. Yemen is also the only country in thesample where lending to HNP went down-in both absolute and relative terms. Part of thisdecline can be attributed to external factors not under Bank control-like wars and other naturalcatastrophes. 4 Though the share of education declined in Zambia, Haiti and Cote d'Ivoire as

' In terms of broad sectors of lending, which this analysis examines, it is not possible to readily identify lending thatwould increase women's access to credit or their participation in the formal labor market.2 Though the sample size is only 12, it is interesting to note that the correlation coefficient is significant at 1 percent.

This is indicated by a negative correlation between the GDI on one hand, and the change in the share of the sixrelevant sectors between the 1980s and the 1990s on the other.4 These were the Gulf War and the Yemen Civil War.

43 Annex III

well, the absolute lending went up. Social Protection gained importance as a category of lendingin the 1990s, and its share rose in all countries from the 1980s level except in Ecuador.5

4. Lending to agnrculture declined as a share of total Bank lending to most countries. InBangladesh, Yemen and Zambia, roughly 80 percent of women in the labor force in the 1990swere employed in agriculture-but both the share of lending and absolute lending to this sectordeclined from the 1980s levels. However, the increased share of lending to agnrculture in Coted'Ivoire, Gambia and Vietnam-employing over 70 percent of the women in the labor force inthe 1 990s-is potentially beneficial to women. Rural transport is another category in whichlending can have special benefits to women. Of the 12 countries in the sample, the share of totallending to rural transport rose in Bangladesh from 1.7 percent to 10.9 percent-leaving thecombined share of agriculture and rural transport practically unchanged between the decades.The share of rural transport lending rose in Ecuador as well, balancing some of the decline inagriculture. In the Philippines, there was a decline in lending to rural transport between thedecades. However, there is no lending to rural transport in countries with high gender disparitieslike Gambia, Yemen, Zambia, Cote d'Ivoire and Haiti.

Table 1: Change in Share of Bank Lending to Specific Sectors between FY 1980-1989and FY 1990-2000

Education HNP Agriculture and 'Relevant'Rural Transport Sectors "

Country GDI' 1980s 1990s 1980s 1990s 1980s 1990s 1980s 1990sBangladesh 0.33 5.1 8.4 2.2 12.6 22.2 22.5 30.9 51.7Cote d'Ivoire 0.34 15.7 15.0 1.4 1.9 9.5 20.0 15.4 42.5Ecuador 0.64 1.6 6.8 0.0 10.3 18.1 15.5 22.8 42.7Gambia 0.28 0.0 34.4 6.5 17.9 11.0 12.2 25.8 85.0Haiti 0.35 7.5 4.8 0.0 10.7 10.3 9.9 25.8 56.9Kyrgyzl11,v 0.65 - 3.4 - - - 31.3 - 46.6

Philippines 0.63 4.2 8.1 1.8 2.0 20.8 20.3 31.3 38.7Poland 0.84 - - - 2.5 - 13.1 - 20.7

SriLanka 0.66 1.3 16.1 1.6 1.6 40.6 11.2 49.8 37.5Vietnam 0.54 - 5.9 - 5.8 - 24.9 - 45.7Yemen 0.31 12.7 10.6 4.0 1.6 29.6 14.4 55.9 48.7Zambia 0.40 5.7 3.4 0.0 2.7 28.7 4.5 38.2 20.2

i. The GDI is taken from the Human Development Report, 1999.ii Includes Education, Social Protection, and Health, Nutrition and Population, Water Supply and Sanitation, Environment,

Agriculture, and Rural transport.iii. There is no lending to these countries in the 1980s.IV. Kyrgyz Republic does not have a GDI value, but, for purposes of representation, it is given a value equal to the average

GDI of the two countries which are closest in terms of the raw data on different components of the GDI.

Ecuador had substantial lending to Social Protection in the late 1980s and the project operated in the 1990s.

44 Annex IV

ANNEX IV: PROJECTS REVIEWED FOR THE EVALUATION,. 1 Lending OED's

COUNTRY SECTOR PROJECT DESCRIPTION Instrument WID Rating ofType Rating2 SAR

Bangladesh Agriculture BWDB Systems Rehabilitation Investment 2 1

Bangladesh Agriculture Tubewell & Low-Lift Pump Investment 2Bangladesh Agriculture Fourth Flood Control & Drainage Investment N/R i

Bangladesh Agriculture Second Small Scale Flood Control Investment 0 0Bangladesh Agriculture Agricultural Support Services Investment 2 2

Bangladesh Agriculture Third Fisheries Investment 2 2Bangladesh Agriculture National Minor Irrigation Development Investment 2 IBangladesh Econ Policy Sixth Technical Assistance Investment 0 0Bangladesh Education General Education Investment 2 3

Bangladesh Electric Power & Other Energy Industrial Energy Efficiency Project Investment 0 0

Bangladesh Electric Power & Other Energy Power Distribution (16 Towns) Investment 0 1Bangladesh Electric Power & Other Energy Third Rural Electrification Investment 1Bangladesh Finance Private Sector Industrial Credit Investment 0 0

Bangladesh Oil & Gas Liquefied Petroleum Gas Transport Investment 2 2Bangladesh Population, Health & Nutrition Fourth Population and Health Investment 2 3Bangladesh Transportation Jamuna Bridge Investment 0 0Bangladesh Transportation Road Rehabilitation and Maintenance Investment N/R 0Bangladesh Transportation Rural Roads and Markets Improvement Investment I IBangladesh Transportation Third Flood Rehabilitation (Emergency) Investment 0 0

Bangladesh Urban Development Urban Development Investment 0 0Bangladesh Industry Jute Sector Adjwatnmnit Adjustinent 0 0Bahgladesh tistsector - ublkc ResourceM M ii ent AdjustmentBangladesb Multisector Second Industrial Sector Credit Adjustment 0 0

Cote d'lvoire Agriculture Forestry Sector Investment 0 0_

Cote d'Ivoire Agriculture Women In Development Investment 2 3Cote d'Ivoire Agriculture National Agricultural Services Investment 2 2Cote d'Ivoire Education Human Resources Management Investment 2 _Cote d'Ivoire Environment Abidjan Environment Protection Investment 0 0Cote d'Ivoire Private Sector Development Privatization Support Investment 0 0Cote d'Ivoire Urban Development Municipal Development Investment 2Cote dIvoire Educaton, HumURews Do cls Adjushnent 2 0

CceOrIqvore i 4utitseett s -~ tie to , hefoi;n Adjuscmenc 0OCote;doIv~ir~ P~b1ic S cc .- Seto 4~i]_ Adjustment 0 0

Cote dl-v&rb Agiltur - i- ltuira[Settr Aajufint Adjustment 0 0|Cote d'voire ultisector conomic Recovery Adjustment N,R U

I OED has used a four point rating system to distinguish levels of gender activities in SARs. In relation to PREM'ssystem, the following applies: ratings of "zero" and "one" are equivalent in both rating systems; ratings of two areequivalent except where OED would pull "best practice" examples and classify them as "three2 See end of table for explanation of WID ratings.

45 Annex IV

Lending OED'sCOUNTRY SECTOR PROJECT DESCRIPTION Instrument WID Rating of

Type Rating SARGambia Agriculture Agricultural Services Investment 2 2Gambia Education Second Education Sector Investment 2 3Gambia Industry Enterprise Development Investment 2 2Gambia Population, Health & Nutrition National Health Development Investment N/R 2Gambia Social Sector Public Works and Capacity Building Investment 2 1Gambia Social Sector Women in Development Investment 2 3Gambia Water Supply & Sanitation Water & Electricity Investment NIR 0

Haiti Agriculture Forestry and Environmental Protection Investment 2 0

Haiti Education Fifth Education Investment 1 IHaiti Electric Power & Other Energy Fifth Power Project Investment 0 0Haiti Multisector Emergency Economic Recovery Investment 0 0Haiti Private Sector Development Industrial Restructuring and Investment 0 0

DevelopmentHaiti Social Sector Economic and Social Fund Investrnent 2 2Haiti Social Sector Employment Generation Investment 2 1Haiti Transportation Seventh Transport Investment N/R 0Haiti Water Supply & Sanitation Port-Au-Prince Water Supply Investment 0 0

Kyrgyz Agriculture RruralFinance Investment 0 0Kyrgyz Agriculture Sheep and Wool Improvement Investment 0 0Kyrgyz Electric Power & Other Energy Power and Distribution Heat Investment 0 0

Kyrgyz Health Health Investment I 2Kyrgyz Social Sector Social Safety Net Investment I IKyrgyz Telecommunications Telecommunications Investment 0 0Kyrgyz AAgriculture Agricultural Private and Enterprise Adjastrment 0 U

Adjustment CreditK yrgyz Finance Financial Sector Adjustment Credit Adjustment 0 U

KrTg) z Mulnsector Rebabilitation Adjustment 0 U

\KTgYZ Public Sector NManagement Pri'.arization and Enterprise Adjustment 0 0K 4Tgyz Public Sector NManagement Public Sector Resource Manangement Adjustment 0 0

_ _ _____________ _ _______ Adjustnent

KNTrgZ Social Sector Social Sector Adjustment Adjustment I _

Philippines Agriculture Rural Finance Investment 0 0Philippines Agriculture Small Coconut Farms Development Investment 0 0Philippines Agriculture Second Rural Finance Investment 2 1Philippines Agriculture Second Communal Irrigation Investment 1 1

Development

Philippines Education Second Vocational Training Investment 2 2Philippines Education Second Elementary Education Investment I 1Philippines Education Engineering and Science Education Investment 0 1Philippines Electric Power & Other Energy Energy Sector Investment 0 0Philippines Electric Power & Other Energy Bacon-Manito Geothermal Power Investment 0 0Philippines Electric Power & Other Energy Manila Power Distribution Investment 0 0Philippines Electric Power & Other Energy Leyte-Cebu Geothermal Investment 0 0

46 Annex IV

Lending OED'sCOUNTRY SECTOR PROJECT DESCRIPTION Instrument WID Rating of

Type Rating SAR

Philippines Electric Power & Other Energy Rural Electrification Investment 0 0Philippines Electric Power & Other Energy Power Transmission & Rehabilitation Investment 0 0

Philippines Finance Industrial Investment Credit Project Investment 0 0Philippines Finance Cottage Enterprise Finance Investment 2 2Philippines Industry Fourth Small and Medium Industrial Investment 0 0

DevelopmentPhilippines Industry Industrial Restructuring Investment 0 0Philippines Multisector Earthquake Reconstruction Investment 0 0Philippines Population, Health & Nutrition Health Development Investment 2 2Phihppines Public Sector Management Tax Computerization Investment 0 0Philippines Telecommunications Telephone System Expansion Investment 0 0Philippines Transportation Second Rural Roads Improvement Investment N/R 0 1Philippines Transportation Subic Bay Freeport Investment 0 01Philippines Urban Development Second Municipal Development Investment 0Philippines Water Supply & Sanitation Water Supply, Sewerage and Sanitation Investment 2 - 1Philippines Water Supply & Sanitation Angat Water Supply Optimization Investment 0 0-Philippines Agriculture Environment and Natural Resources Adjustment I I

ManagementPhilippines Multisector Economic Integration Adjustment I 1

Poland Agriculture Agricultural Development Investment 0 0Poland Agriculture Agroindustries Export Development Investment 0 0Poland Agriculture Forest Development Support Investment 0 0Poland Electric Power & Other Energy Heat Supply Restructuring Investment 0 0Poland Environment Environment Management Investment 0 0Poland Finance Industrial Export Development Investment 0 0Poland Oil & Gas Energy Resource Development Investment 0 0Poland Public Sector Management Private Enterprise Development Investment 0 0Poland Social Sector Emnployment Promotion Investment 2 1Poland Telecommunications Telecommunications Investment 0 0Poland Transportation First Transport Investment 0 0Poland Agriculture Agricultural Sector Adjustment Adjustment 0 0Poland Finance Financial Institution Development Adjustment 0 0Polnd ' Nulisector tEMS'AL Adjustmient 0 0Polana mitltisector Debt and Debt.Service Reduction Adjustment 0 0

Sri Lanka Agriculture Agricultural Research Investment N/R _ 0Sri Lanka Agriculture Second Smallholder Rubber Investment 1 1

RehabilitationSri Lanka Agriculture Forest Sector Development Investment 0 0Sri Lanka Agriculture National Irrigation Rehab Investment 0 1Sri Lanka Agriculture Agricultural Extension 2 Investment 2 1Sri Lanka Education Vocational Training 2 Investment N/R 0Sri Lanka Education General Education Investment 1 1Sri Lanka Electric Power & Other Energy Distribution & Transrnission Investment 0 0Sri Lanka Electric Power & Other Energy Power Distrib.& Transm. 2 Investment 1 0Sri Lanka Finance Industrial Development 3 Investment 0 0

47 Annex IV

Lending OED'sCOUNTRY SECTOR PROJECT DESCRIPTION Instrument WID Rating of

Type Rating SAR

Sri Lanka Finance Private Financial Development Investment 0 0Sri Lanka Population, Health & Nutrition Health & Family Planning Investment 2 2Sri Lanka Public Sector Management Small & Medium Industries 4 Investment 1 1

Sri Lanka Social Sector Poverty Alleviation Investment 2 3Sri Lanka Telecommunications Telecommunications 2 Investment 0 0Sri Lanka Transportation Roads 3 Investment 0 0Sri Lanka Transportation Colombo Urban Transport Investment 0 0Sri Lanka Urban Development Municipal Management Investment N/R 0Sri Lanka Water Supply & Sanitation Water Supply & Sanit. Rehab. Investment N/R 0Sri Lanka Water Supply & Sanitation Community Water Supply and Sanitation Investment 2 3

Sri Lankh bNluriseccor Econ. Restructuring Adjustment 2Sri Lank-a Public Sector Mlanagement Public Mfg. Enter. Adj. Adjustment I 0

Vietnam Agriculture Agric. Rehabilitation Investment 0 0Vietnam Agriculture Rural Finance (ends Sept 2001) Investment 0 2Vietnam Electric Power & Other Energy PowerDevelopment Investment 1 0Vietnam Education Primary Education(ends June 2002) Investment 0 2

Vietnam Energy Power Sector Rehab. and Expansion Investment 0 0

Vietnam Finance Banking System Modernization Investment 0 0

Vietnam Health National Health Support(ends Sept Investment 2 32003)

Vietnam Health Population and Family Health Investment 2 3(ends June 2003)

V ietnamu_ _tisector Stractural Adjustment Adjustment 0 O\Viemam Finance Debt & Debt Service Reduction Adjustnent 0 0

Yemen, Rep Agriculture S. Regional Agricultural Dev. Investment N/R 3Yemen, Rep Agriculture Northern Regional Agricultural Investment 2 2

DevelopmentYemen, Rep Agriculture Tihama Regional Agri. Dev. 5 Investment N/R 2Yemen, Rep Agriculture Eastem Region Agri. Deve. Investment 2 1Yemen, Rep Agriculture Wadi Hadramawt Agr. Dev. 3 Investment 0 1Yemen, Rep Agriculture Fisheries Development 4 Investment 2 3Yemen, Rep Infrastructure & Urban Emergency Flood Reconstruction Investment 0 0

Development

Yemen, Rep Education Teacher Training Investment N/R 3

Yemen, Rep Electric Power & Other Energy Power 3 Investment 0 0Yemen, Rep Population, Health & Nutrition Health Sector Development Investment 2 2Yemen, Rep Social Protection Emergency Recovery Investment 2 1Yemen, Rep Population, Health & Nutrition Health Development 2 Investment 2 2Yemen, Rep Urban Development Taiz Flood Disaster Prevention Investment 0 0Yemen, Rep Water Supply & Sanitation Greater Aden Water Supply 2 Investment N/R 0Yemen, Rep Water Supply & Sanitation Tarim Water Supply Investment 0 0Yemen, Rep Water Supply & Sanitation Al Mukalla Water Supply Investment 0 1

48 Annex IV

Lending OED'sCOUNTRY SECTOR PROJECT DESCRIPTION Instrument WID Rating of

.______ _ . _Type Rating SARYec*Jstep M tisecror Economic Recovery Adjustment 0 0

Zambia Agriculture Agri. Research & Extension Investment N/R IZambia Agriculture Coffee 2 Investment N/RZambia Agriculture Agricultural Marketing and Processing Investment 2 2

Zambia Education Education Rehabilitation Investment 2 3Zambia Mining Mining Technical Assistance Investment 0 _ )

Zambia Public Sector Management Technical Assistance 2 Investment N/R (IZambia Public Sector Management Privatization & Industrial Reform TA Investment 0 0

(PIRC)

Zambia Social Sector Social Recovery Investment 2Zambia Transportation Transport Engineering and Technical Investment 0 0

Assistance (TETAP)

w7 7 s; .... t- . -; ce . -..........P.v*.,& Indus:-Eeform 2 Adjustment 0 - 0Pdivai B: Iaii .Reformn A eA 2Economic &Ucia Adj. _d_ ______

i ->. F~~tor - Econ0iC an&SMWiarAdj 2 ji m 0 02_____}. ___ - hy,iMo. Sector Management Economic Recovery & Investment Adjustment 0 0

180 projectstotal

Starting in 1988, a systematic women in development (WID) rating system was instituted for all World Bank projects. The ratingsystem is based on appraisal documents and the rating process is carried out and maintained by the Poverty Reduction andEconomic Management network (PREM). Three possible ratings exist under this system:

0 = projects with no attention to WID/genderI = projects with some discussion of WID/gender issues but with no specific action to address them2 = projects with concrete, specific activities addressing WID/gender- related issues.N/R = Not Rated

OED's evaluation of project appraisal documents used a rating system which ranges from 0 to 3. The rating is the average of theratings of three questions: a) Does the appraisal document display an understanding of gender issues/disparities? b) Does theappraisal document include specific interventions or components to assist women? c) Does the appraisal document display anunderstanding of how the project will impact women? Answers to these questions were put into the followingratings:

0 = no reference 1 = poor 2 = satisfactory 3 = highly satisfactory

49 Annex V

ANNEX V: SOCIOECONOMIC INDICATORS USED FOR THE EVALUATIONBaseline Indicators for Six Sample Countries (Source: WDI)_____ ___________ _________________

Bangladesh Yemen Zambia80-84 85-89 190-94 195-98 80-84 j85.89 j90-94 j95-98 80-84 185-89 190-94 195-98

iSclurce: IIDR 1915 &1990) Rl199h IDR 1999 l{DRI1995 Rl9lFDR1995 11DR199'lk,es.der Dle%eloprnent ludex iGiDii ii 3U ' 43 3 fI '.1 i 4L 43;],ndci Empn%ermeru Mlea.ure l'iFNl I ( .. 3 27 .31Sh.ar -i nt31ea in narional leeixiarnre 4,. s%onien IF IIIt, Ii I.I) 6 71 I 1'

EducationSchool enrollment, primary, female (% gross) F 50.08 55.84 66.30 ... . 44.90 39.90 86.64 98.05 87.70 85.60School enrollment, primary, male (% groas) M 73.70 72.96 76.50 ___... . 112.90 99.90 100084 108.73 93.80 91.40Primary School enrollment ratio, female to male G 0.68 0.77 0.87 0.40 0.40 0.86 0.90 0.93 0.94School enrollment, secondary, female (% gross) F 8.98 11.68 12.70 8.50 14.30 11.60 15.13 21.10School enrollment, secondsry ,male (% grass M 25.90 25.54 25.10 ___ 36.40 53.20 22.64 26.68 33.60Secondary School enrollment ratio, female to male G 0.35 0.46 0.1 'i ____023 0 7 0.52 0.57 0.63 ____

Primary education, teachers (% female) F 7.83 11 98 19.38 ___ 16.60 40.66 44.06. 44.24 43.33Secondary education, teachers (% female) F 6.69 9.45 9.55 . . 21.141Health ____

Fertility rate, total (births per woman) F 6.06 4.90 3.86 -3.2( 7.84 7.70 7.37 6.39 6.96 6.5 6.20 554Birtha attended by health staffo/% of total) F 2.00 5.00 8.50 8... 16.00 43.00 .. 38.0( 42.67 47.00Under 5 mortality ratio (per 1 000 live births) N 192.00 139.00 126.00 100.0 198.00 . 130.00 101.00 149.00 191.0( 195.50 190.50Life expectancy at birth, female (years) F 48.66 53.51 55.18 58.39 50 10 52.00 52.80 55.76 52.12 50.6( 49.88 42.91Life expectancy at birth, male (years) M 49.50 53 55.14 58.27 4820 51.00 52.04 54092 49.62 48.7 48.14 42.81Life expectancy at birth ratio, female to male G 0.98 1.01 1.00 1.0 1 04 102 1.01 1.02 1.051 14 1.04 1.00

Economaic I____

Arable land (hectares per person) N 0.10 0.09 0.07 0.06 0.15 0.13 0.10 0.09 0.84 0.75 0.64 0.57Forest area as percent of total land area N 8.00 7.00 __ _0.00 0.00 ___ 44.00 42.00Employees, agriculture, female (% econ active pop) F 80.90 84.90 77.50 98 30 87.80 84.70 82.80Employees, agriculture, male (% econ active pop) M 66.50 54.20 54.40 60.10 49.70 _____ 69.00 ___ 67.90 ____

Ratio of female to male emiployees, agriculture(% econi active pop) G 1.22 . 1.57 1.42 1.64 ___. 1.77 ___.. 1.23 ____. 1.22 ____

Emiployees, industry, female (% econ active pop) F 13.90 8.80 7.60 0890 5.50 . 2.80 .. 3.20Employees, industry, male (% econ active pop) M 4.90 1560 10.80 19.20 ____ 21.60 . 12.50 ____. 12.80 ____

Ratio of female to miale emiployees, industry(% econ active pop) G 2.84 0.56 0.70 0.04 __ _ 0.25 ____. 0.22 ____. 0.25 ____

Employees, services, female (% econ active pop) F 5.30 2.10 11.00 -090 6.70 . 12.60 .. 14.10Em~ployees, services, male (% econ active pop) M 28.60 25,50 33.70 20.70 ___ 28.80 . 18.50 ___.. 19.301____Ratio of female to malceemployees,saervices(% econ active pop) G 0.19 0.08 0.33 0.04 0.__ 23 ____. 0.68 .. 0.73Safe water (% of population with access) N ____ 40.00 84.20 - __ _ 39.00 ___.. 48.00 43.00Safe water, rural (% of rural population with access) N 40.08 54.50 82.50 ______ _ 14.00 .. 32.00 27.00Safe water, urban (% of urban population wi th access) N 24.00 27.00 44.50 74.00 ___ 70.00 64.00 ____

Sanitation (% of population with access) N .. 4.00- 35.00 ___ 19 00 ___ 47.00 23.00-Illiteracy rate, adult female (% of females 15.-) F 81.84 79.00 75 72 72.50 93.38 90.20 84.90 79.35 50.98 44.90 30.64 32.88filiteracyrate, adultmale (% ofmales 15+) M 8.0 555 52:72 49.80 58.12 50.36 41.80 35.98 27.10 23.60 20.02 16.97Illiteracy rate, ratio femnale to male(age 15+) G 1.41 1.421 1.44 1.46 1.61 1L79 2.03 2.21 1.88 1.91 1.93 1.94

Rural Roads per capita (kin) N

Institutional"0

7 7 7t_______ _____ _____ _____ _____ ____

Signed CEEDAW (N, Y. Y w/res) Yes Yes YesNational machiner to implemenet poicy Yes Yes Yes

F:female; M: mnale; G:gender; N: neutral

All data in each period is an average of the data available fur Chose years

50 Annex V

Baseline Indicators for Six Sample Countries (Source: WDI)__________________Ecuador ______RpublicVietnam

____________________________________________________________________ 80-84 .85-89 1 90-94 195-98 80-84 85-89 190_9 95--98 80-84 85-89 190-94 95-98Gencral--

re~~~~~.-,[[DR 1"S WOM.' 195 199 9l1U9511)199 DRI993 BER199G' 0.6 0.73 - 4.01.5 -0.66

IMM" ~~~~~~~~~a0. 0.52ilshu ofag ~ hr@ women) F .20 tTAO .. - 26.20

________________________________________________________ N I Sbti9s Ž84 45 2.891 IC 3I60 V; - 'is~l2 tM4 67 2.2.J05I '"KX'4111 I i' 34 1.605 fl

Schoolenrollment, primary, femnale (%/gross) F 117.70 118.97 12167 120.30 114.30 11517 112.12 102.70 103.60 99.90 111.40School enrollment, primary, male (% gross) M 119.96 120.60. 122.17 128.80 117.20 112.70 110.72 105.50 108.45 105.90 115.50Primary School enrollment ratio, female to male Gi 0.98 0.99 1.00 0.93 0.98 1.02 1.01 0.97 0.96 0.94 __ _ 0.96School enrollment, secondary, female (% gross) F 56.22 59.10 51.47 - 108.20 103.63 96.1,6 83.00 39.70 39.40 33.44 -46.00School enrollment, secondary, male (%/ gross) M 55.24 56.80 50.50 _____ 111.70 106.57 92.52 74.60 44.30 42.54 35.00 48.00Secondr School enrollmnent ratio, female to male Gi 1.02 1.04 1.01 0,97 0.97. 104 1.11 0.90, 0.93 0.96 0.96Primary education, teachers (% female) F 64.98 65.30 66.64 67.65 88.18 81.09 8137 8.6 68.15 70.33 ___ 77.41Secondary education, teachers (% female) F 38.87 41.04 ---- ---

Fertilit mtle, total (irths per woman) F 4.84 5.19 362 2.98 407 402 3.48 2.97 4.77 38 343 2Births attended by health staff (%of total) F 5.0 4.5 6. 64 ____ ____. 9.00 9.0 8. 79.0Under 5mortality ratio (perI loo0live births) N 101.00 70.00 51.00 38.0 . 46.90 41.66 40.19 105.00 - 52.0 43.50Life expectancy at birth, female (years) F 65.92 54.41 71.02 72.93 70.10 70.31 71.98 71. 65.56 67.94 69.10 70.72Life expertancy at birth, male (years) M 1.94 52.381 66.06 6.1 6. 62.451 63.50 62.3 6.7 6.7 46 65.94Life expectancy at birth ratio, female to male G 1.06 1.04 1.08 1.08 1.15 1.131 1.13 1.14 1.06 1.07 1.07 1.07Econonmic Arable land (hectars pe eson) N 0.19 0.17 0.15 0.13 .____. 0.29 0.29 0.11 0.09 0.08 0.08Forest area as percent of total land area N . .. 44.00 40.00 .. ____. 4.00 4.00 .. ____. 30.00 28.00Employees, agricltwre, femnale (% econ active pop) F 21.90 .. 1.68 1.93 32.90 ... 48.50 75.30 .. 73.10 71.10Emiployees, agriculture, male (% eron active pop) M 44.30 ____. 10.28 9.371 34.50 _ __... 47.60 71.20 .. 69.50 70.20Ratio of f:rale to miale employees, agrculture(% econ active pop G 0.49 ____. 0.16 0.21 0.95 _ __... 1.02 1.06 _ _.. 1.05 1.01Employees, industry, female (% econ ac-tive pop) F 15.40 .. 17.98 14.73 23.40 ... 7.20 10.00 .. 10.80 8.60Enmployees, industry, male (%/ econ active pop) M 21.40 ____. 28.90 26.40 33.70 ... 12.30 16.20 ___.. 17.10. 12.30Ratio of female to miale employees, industry(% econ active pop) G .20.62 0.56 0.69 0.59 0.62 0.63 0.70Emiployees, services, female (% econ active pop) F 62.70 8034 83.33 43.80 38.10 14.70 .. 16.00 20.20Einiployees, services, male (% econ active pop) M 34.20, 60.78 64.20 31.50 31.40 12.601___. 13.40 17.501Ratio of female to male employees, services(% econ active pop) G1 .8113 130 1381.21 1.171___. 1.19 1.15Safe water (% of population with access) N_ _______70_00 80.50 . ___... 36.00Safe water, rural (% ofnrural population with access) 4240 00.. .. 42A 32.00Safe water, urban (% of urban population with access) N_ 79_00 82_00 93___0 53.00Sanitation (% of population with access) N_ 57_00 64.00 ___. 21.00 ___

Illiteracy rate, adult female (% of females 15+) F 20.38 16.92 14.12 11.93 17.76 14.62 12.12 tO OSIlliteracy rate, adult male (% of males 15+) M 13.72 11.341 9.40 7.95 _ _ 64.44 5.82 5.30 4.85Illiteracy rate, ratio female to male(age 15+) Gi 1.49 1.49 I 50, 1.50 ___ _2.76 2.51 229 2.07Rural Roads per capita (Ian)

1n tt tio aI E ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _

SindCEEDAW (N, Y, Y w/res) Yes Yes YesINational machinery to inmplement policy Yes Yes Yes

F:female, M: male; G:gender; N: neutral

All data in each period is an average of the data available for those years

51 Annex V

Baseline Indicators for Six Sample Countries (Source: WVDI) ________________

Cote d'Ivoire Gambia Haiti-J80-84 185-89 190-94 j95-98 80-84 j85-89 90-94 j95-98 80-84 185-89 J90-94 195-98

General I_ _ _ _ _ _ _ _ _ _

(14ource: FIDR 1995 & I99) RLDR 1995 HD R 1999 fIDR9 R 119 [ I999 DfR1995 TDR1999Vjendei Dceleopmsenl Inde% (GDII (G 1).34 (1.40 0.28s ( 39 515 0 43Genler Ernpu%errmeni: Mcasure 1GEM i 3 0) lb( 0 3 2 0.24 O 35Sh.re ofw&&I in fl3tioflal lIC slatizet 5. %omrniii f 4 '9) S 64). '8 2.01) 11M'liDP !'c' C~l U PPP 'sj Ni 12Ž3 4 1,14. 1 I All 4. I' I W~ 1 '53 I 412 ", 1,4'.')1 52 I 1`4 1i. I '4 ii 1 I I .10.'!EducationSchool enrollment, primary, femtale (% gross) .F 61.18, 58.30 55.70 59.45 45.70 50.84 55.96 67.20 77.55 61.32 46.40School enrollment, primary, male (% gross) M 88.82 82.48 77.40 80.75 79.70 78.58 79.36 87.10 88.75 66.40 49.201____Pfimary School enrollment ratio, female to male G 0.69 0.71 0.72 0.74 0.571 0.65 0.71 0.77 0.87 0.92 0.94 ___

School enrollment, secondary, female (% gross) F 11.30 12.40 15.10 15.73 8.80 9.44 14.86 18.80 16.25 19.22 20.40School enrollment, secondary, male (% gross) M 26.65 27.501 30.34 32.40, 20.55 22.02 28.30 30.40 17.53 20.73 21.40Secondary School enrollment ratio, female to male G 0.42 0.45 0.50 0.491 0.43 0.43 0.53 0.62 0.93 0.93 0.95 ____

Primary education, teachers (% female) F 17.66 18.62 19.11 20.501 31.88 32.77 30.92 28.99 51.08. 46.49 44.57Secondary education, teachers (% female) F . . 25.05 23.25 17.00 16.74HealtbFertility rate, total (births per woman) F 7.41 690 5.94 5.04 6.50 6.20 5.95 5.4 5.80 5.60 494 4.32Births attenided by health staff(% of total) F 13.00 40.00 45.00 45.00 41.00 54.0 ___ 44.0. 27.00 33.33 .. 20.50Under 5 mortality ratio (per I1000 livehbirths) N . .. 150.00 162.00 127.00. .. 131.00 120.50Life expectancy at birth, Female (years) F 52.00 5250 50.66 4700 42.60 49.00 52.20 55.00 52.80 54.20 55.24 56.10Life expectancy at birth, male (years) M 48.40 4950 48.22 45.851 39.401 45.10 47.98 51.44 49.90 5i.00 51.40 51.33Life expectancy at birth ratio, female to male G 1.07 106 I 05 1.03 1.08 1.0 1.09 1.07 1.06 1.06 1.07 1.09EconomiceArabic land (hectares per person) N 0.23 0.23 0.21 0.21 0~26 0.22 0.17. 0.16 0.10 0.09 0.081 0.08Forest area as percent of total land area N ____ 18 00 17.00 . .. 10.00 9.00 . . 1.00 0.00Employees, agriculture, female (% econ active pop) F 72.00 . .. 91.80 .. 53.10 49.60Employees, agriculture, male (% econ active pop) M 54.201. . 74.00 ___.. 81.30 76.10 _______

Ratio of femnale to male employees, agriculture(% econ active pop) G ____ 1.33 ____. . 1.24 .. 0.65 0.65Employees, industry, female (% econ active pop) F 5.60 . .. 2.40 .. 7.90 9.30Employees, industry, male (% econ active pop) M ___ 11.50 -___ ____.___. 11.60 .. 8.20 8.60 _______

Ratio of femnale to male employees, industry(% econ active pop) G ____0.49 . .. 0.21 ___.. 0.96 1.08 _______

Emnployees, services, female (% econ active pop) F 22.40 . .. 5.80 .. 39.00 37.80Em..ployees, services, male (% econ active pop) M ____34.40 ____ ___. . 14.30 .. 10.50 12.90 ____ ___

Ratio of female to male employees, scrvices(% econ active pop) G ____0.65 . .. 0.41 .. 3.71 2.93 _______

Safe water (% of population with access) N 20.00 72.00 41.60 45.00 70.00 ____.. .. 28.00Safe atr, rural (% ofnrural population with access) N 10.00 81.00 .. 27.00 33.00 56.00 ___ ___.. 23.00 ___

Safe water, urban (% of urbasn population with access) N 30.00 59.00 .. 100.00 100.00 92.00 ____ . 37.00Sanitation (% of population with access) N 17. 10o 50-00 54.00 .. 37.00 ___ ___ 20.50 24.00 ___

IlliteracyrTate, adult femalce(%of females 15-i) F 84'73 80-20 73.34 66.65 87.17 8336 78.32 74-07 70.28 65.96 60.86 56.00Illiteracyrate, adult male (% ofmales 15+) M 63.97 59.72. 53.92 48.82 76.58 71.70 65.54 59.97 63.65 59.90 55.48 51.33Illiteracy rate, ratio female to malc(age 15+) G 1.33 14 1.36 1.37 1.14 1.16 1.19 1.24 1.tO 1.10 1.10l 1.09Rural Roads per capita (kin) NInstitutional-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Siped CEEDAW (N. Y. Y w/rrs) YsYes YesNational machinery to implemenrt policy Yes Yes N/A

F:fermale; M: male; G:geader; N: neutral

All data in each period Is an average of the data available for thote years

52 Annex V

Baseline Inidicators for Six Sample Countries (Source: WDI)________________Phlippinles Poland Sri Lanka

_________ ________ _________ ________ ________ _________ ________ _______ 80-84 j85-89 190-94 195-98 80-85 11~-59I8Q14 5 80-84 [85 9 190-94 [9s-98

General- - - _ __

(Eaure; EIDR I995FI

999b HDRI99S 999U HDR1995 111R99 IIDR1999

Gender M8e~JtwfamVP40% (OD - 043 0.74 .84 0,80 0.6 0.71Genile Bmpo 0.44 04 0.43 O.50 0.2 0.32

Share fmats irw%oine ) F 11.20 12 90 F 1300 12.90 490 5.3016DP per-ipita i.PPP Si N - J3 43 >t~'I 3,1i.,26 3)l lla l IV .4 I,l,'us I fot' 14 ',25 14 'I tV

School enrollment, primiary, fem-ale (% gross) F 108.10 108,721 109.50 113.10 100.03 99.66 97.64 95.50 102.23 103.34 10666 110.25School enrollment, prinmary, male (% gross) M 110.25 1109 110.85 115.10 101.33 100.76 99.26 97.30 105.70 105.72 108.90 112.50Primr School enrollment ratio, femtale to male G 0.98 0.98 0.99 0.98 .. 0.97 0.98 0.98 0.98School enrollment, secondaTy, femnale (%/ gross) F 67.65 68.16 75.86 77.85 79.40 82.36 90.92 97.40 61.63 71.90 77.74 78.30School enrollment, secondary, miale (% gross) M 64.10 69,22 76.12 77.00 73.75 77.86 88.24 97.80 55.05 65.44 70.80 71.50SecondaTy Schsool enrollment ratio, fenmale to male G3 1.06 098 1.00 1.01 1.08 1.061 103 1.00 1.12 1.10 1.10 1.10PrimaTy education, teachers (% female) F 88 41 -- --------- ____.. 80.98 89.50Secondary education, teachers (% female) F 95.07 -19 20Health _____________________________ 19 20

Fertility rate, total (births per woman) F 4.74 4.3 40 3.65 2.34 2.18 193 1.52 3.25 2.60 2.50 2.15.Births attended by health staff (% of total) F 57.00 81.67 76.10 53 ___0 _ 98.85 98.0 87.00 90,45 94.0 _______

Under 5 mortality ratio (per 1000 live births) N .. 72.0 62.0 44.33 .. 21.80 19.48 M13. . 23.0 18.50Life expectancy at birth, female (years) F 63.70 65.9 67.74 70.3 75.15 75.25 75.72 76.83 71.50 73.0 73.96 75.48Life expectancy at birth, miale (years) M 60.20 62.2 64.0. 66.6 67.02 66.821 66.84 68.27. 67.00 68.50 69.4 71.01_Life expectancy at birth ratio, female to male 13 1.06 1.0 1.0 6 L1.2 1.13 1.13 t.13 1.07 t,07 L1. 1.6

Arabic land (hectares per person) N 0.09 0.08 0.07 0.07 0.40 0.38 0.37 0.37 0.05 0.05 0.05 0.05Forest area as percent of total land area N -___ ___ 27.00 23.00 ___. .. 28.00 29.00 ____. ___. 29.00 28.00Employees, agriculture, femiale (% econ active pop) F 31.26 29.93 . .. 24.90 21.60 51.00 .. 44.55 40.85Empqloyees, agricultur, miale (% econ active pop) m ____ 53.32 49.23 ___. .. 24.651 21.87 44.20 ___.. 38.851 34.15Ratio of female to miale employees, agriculture(% econ active pop) G1 ---------- __ ____ ____ 1.01 0.99 1.15 ____. 1.15 1.20Employees, industry, femiale (% econ active pop) F . .. 13.42 12.80 . .. 20.95 20.67 17.50 .. 24.00 26.45Employees, industry, male (% econ active pop) M ___.. . 16.96 18.40 . .. 40.90 41.03 1890 .. 17.90 21.25Ratioof female tomnale effloyees, indutry% econ active pop G ____. . . . __. . 0.51 0.50 0.93 _ _.. 1.34 1.24Employees, services, fenale (% econ active pop) F . .. 55.28 57.23 . .. 54.20 57.70 27.60 .. 29.25 30.80Eniployees, services, miale (% econ active pop) M I . . 29.62 32.33 . .. 34.80 37.07 29.90 ___.. 38.45 38.65Ratio of female to ale enployees, services(% econ active pop) G . .. 1.87 1.77 . . 1.56 1.56 0.92 ____. 0.76 0.80Safe water (% ofpopulation with access) N ____. 68.70 81.35 83.00 ___. 81.80 . . 37.00 46.001Safe water, rural (% of rural population with access) N . 67.70 77.00 81.00 .. 73 00 ___ ___ 26.00 47.00 _ __

Safe water, urban (% of urban population with access) N ___. 80.50 93.00 91.00 _ _.. 89.00 ---------___ 76.00 43.00Sanitation (% ofpopulation with access) N ____. 63.35 77.00 77.00 .____ 100.00 _ __.. 52.00 _ __

Illiteracy rate, adult feniale (%/of females 1St) F 10.83 9.10 7.32 5.83 0.95 0.64 0.42 0.33 19.05 16.72 14.38 12.38Illiteracy rate, adult male (% ofmales 15+) M 9.10 7.78 6.44 5.27 0.55 0.42 0.32 0.30 8.50 7.66 6.82 6.10Illiteracy rate, ratio femiale tomnale(age 15+) G 1.19 1.17 1.14. 1.10 1.73 1.52 1.31 1.08 2.24 2.18. 2.11 2.03Rural Roads per capita (Ian) NIlstitutional**_ _ _ _ _ _ __ _ _ _ _ _ __ _ _ _ _ _ _ _ _

Signed CEEDAW (N, Y, Y wires) Yes Yesi YesNational machinery to implemenlt policy Yes Yesl Yes

F:female; M: male; G:gender; N: neutral

All data in each period is an average of the data available for those years

53 Annex VI

ANNEX VI: NATIONAL WID POLICIES IN TWELVE COUNTRIES

1. This annex summarizes the WID/Gender policies and action plans of the 12 countries in theevaluation. Each country, to differing degrees, provides an entry point for addressing gender/WIDissues.

2. Bangladesh has had a policy for the advancement of women since 1976. The Ministry ofWomen and Children Affairs serves as the focal point in the national machinery for the advancementof women's issues. Its role is to facilitate government-wide mainstreaming of a gender equalityperspective in all policy areas. It advocates, coordinates, communicates and monitors implementationof the National Action Plan (NAP). The principal goals of the NAP are: to remove legal, economic,political or cultural barriers, to raise public awareness about women's differential needs, to improvewomen's development, and to provide full equality of opportunity.

3. In Cote d'Ivoire, the Ministry for the Advancement of Women was mandated with theresponsibility to outreach services to women in rural areas. Cote d'Ivoire has seen tumultuousevents, including a coup and cancelled elections in December 2000. Cote D'Ivoire has no nationalpopulation plan no information is available of any Government National Policy for women.However, it appears that the Ministry for the Advancemnt of Women has been changed to theMinistry of Family, Women and Children, now headed by Henriette Lagou.

4. During the last decade, Ecuador has witnessed a major advance in institutional andlegal reforms related to gender. The first National Office for Women was established in 1970 but itwas only during the 1980s when this office evolved into an independent body. In October 1997 theGovernment created the National Council for Women (CONAMU) and placed it directly under theresponsibility of the Presidency. The objective of CONAMU has been to institutionalize public sectorpolicies for women's development and gender equity and to promote women's participation in thedevelopment process. CONAMU has a strong relationship with Ecuadorian civil society. Under itscurrent structure, CONAMU enjoys a certain level of budgetary and administrative autonomy. Oneof CONAMU's main achievements is SIMUJER, a gender disaggregated data base which compilesstatistics from numerous sources in the area of demographics, health, education, violence, politicalrepresentation, etc. This data base is the first of its kind in the region.1

5. In 1980 the Gambia government established the National Women's Council and theWomen's Bureau. The goals of these institutions were to address women's issues. Their efforts led tothe adoption in 1987 of a two-pronged WID strategy (1) to improve women's economic status, and(2) to increase women's access to basic social services.

6. In 1994, the Haitian Government established a Ministry of Women's Affairs to promotethe full participation of women in the development process and to ensure the integration of genderissues into all Government plans and policies. The priority areas for enlisting a more activeparticipation of women have been identified in a 5-year plan called "Offensive 2001". These include(i) revision of gender-related discriminatory laws and the enactment of legal provisions to protect

'World Bank 2000f, pg.9.

54 Annex VI

women against all forms of violence and abuse; (ii) promotion of a more gender balanced partnershipat all levels of decision-making; (iii) eradication of the burden of poverty on women by providingthem access to productive activities and resources; and (iv) promotion of public campaigns to raiseawareness on the condition of women.

7. The president of Kyrgyz Republic declared 1996 to be "the Year of the Woman" andinitiated a national program, AYALZAT, to address the negative aspects of the economic transitionfor women. AYALZAT's objectives are to i) promote equal rights between men and women, ii)address the high social costs of transition on women, iii) provide women with equal access todevelopment opportunities within the emerging market economy, and iv) provide women's full andactive participation in political and social life. The president appointed a special State Committee onFamily and Women under his office to cooperate with the Deputy Prime Minister, who is responsiblefor social policy. The AYALZAT committee helps to coordinate line ministries, local governments,the non-governmental sector and donors to oversee the integration of women's concerns throughouttheir programs and projects and to develop implementation plans2.

8. Philippines has a clear gender equality policy, an established machinery for supportingits implementation, and a government regulation that requires Government agencies to spend at leastfive percent of their budget for gender mainstreaming activities. The National Commission on theRole of Filipino Women is the agency in charge of monitoring the implementation of gender policyby government agencies. It ensures the gender responsiveness of national development plans;coordinates the preparation, assessment, and updating of the National Plan for Women; and ensuresthe implementation of the Plan at all levels; undertakes advocacy to promote economic, social andpolitical empowerment of women; provides technical assistance in the setting-up and strengtheningof mechanisms on gender mainstreaming. Furthermore, it can issue orders, circulars and guidelines.It is also in charge of coordinating the various organizations involved in women's affairs.

9. Poland has no official WID or gender policy. However it had, until 1997, a nationalinstitution in charge of advancing women's issues. The Government Plenipotentiary for Women andFamily was in charge of analyzing women's social situation, supporting women's organizations,cooperating with international organizations and securing the execution of international obligationsas written in ratified conventions and documents. In 1997 it was renamed the Plenipotentiary forFamily Affairs. Its mandate does not include working for the advancement of women and insteadadvices govermment mainly on matters relating to the family and children. The only group workingon behalf of women's equality in the Sejm (House of Representatives) is the Parliamentary Group ofWomen, a voluntary caucus of Parliamentary members.

10. In Sri Lanka a state ministry for women's affairs was established in 1983. Provincialministries were consequently set up. In the early nineties, a Women's Charter ensured statecommitment to "the full development and advancement of women, ... on a basis of equality withmen". A National Committee on Women was set up to monitor the progress with regard toimplementation of the Women's Charter, to receive complaints of gender discrimination and otherissues and problems of women, and to help formulate strategies to solve them. More recently the

2 Bauer, Armin, David Green, Kathleen Kuehnast. 1997, pp.74 ,7 5.

55 Annex VI

Ministry released a National Action Plan for Women that focuses on problems and issues of criticaland specific concern to women in the country and sets out strategies and activities for resolving them.

11. In Vietnam, the National Committee for the Advancement of Women (NCAW) wasestablished in 1993. The NCAW is responsible for recommending to the government ways andmeans of implementing policies related to women, inter-agency coordination, and cooperation withthe UN and other international organizations. Vietnam's National Plan of Action for theAdvancement of Women has 11 specific objectives, including increasing women's income,improving women's health services, education and training.

12. Although Yemen's constitution declares equal rights and obligations for men and women,and makes discrimination on the basis of sex illegal, the government's capacity to enforce such lawsis weak, mostly due to inadequacy of the administrative apparatus. New laws introduced sinceYemeni unification provide women more security rights, yet without effective enforcement, these arevirtually ineffectual. In 1997 the government adopted the Yemeni Women's National Strategy.Among the institutions established to work on women's issues is the Women National Committee.It's duties include; cooperating with local, regional and international organizations involved inwomen's projects, conducting studies relating to women, contributing to women's legal awareness,and holding workshops and conferences. Its effectiveness is hampered, however, by lack of adequatefunding; absence of coordination at a national level; unavailability of data disaggregated by gender;and ambiguity of its goals.

13. In Zambia, the responsibility to coordinate and monitor WID/Gender policy implementation isleft to the Gender and Development Department within the President's Cabinet. With support fromNORAD and other donors, the Government adopted a Gender Policy in 1999, which is unique for itsconceptual clarity, and goes beyond the Beijing Platform for action in focusing on the welfare of bothmen and women.

56 Annex VII

ANNEX VII: RESULTS OF STAFF SURVEY

1. The Operations Evaluation Department (OED) administered, via email, a survey to non-administrative bank staff, including field personnel, on December 2000. OED received 391completed surveys. Of the respondents, 40 percent were women. This response rate is slightlyhigher than the percent of female high-level staff members in the Bank. According to BusinessWarehouse December figures, women occupy 33 percent of positions with a grade of F or higher. Asummary of the responses follows.

"Areyou aware that the Bank has a gender policy?"92 percent of respondents are aware that the Bank has a policy on gender. There were no statisticallysignificant differences between men and women regarding their awareness of a gender policy. Yearsin the Bank is positively related to awareness, meaning that the longer a staff member has worked inthe Bank, the more likely he/she is to be aware that the Bank has a policy on gender. Task teamleaders, as well as managers, are also more likely than other staff members to be aware of a Bankpolicy on gender.

"Have you heard of OP4.20?"72 percent responded yes. Here again there were no statistically significant differences between menand women. Although task team leaders are more likely than other staff members to have heard ofOP4.20, managers are not.

"Have you read OP4.20?"42 percent answered yes. Again, there are differences based on gender. Interestingly, years in theBank is also not a predictor of whether staff members have read OP4.20. Although managers aremore likely than other staff members to have read the policy, task team leaders are not.'

Do you address gender!WID issues in you operational work?52 percent answered yes, while about one third (31 percent) answered "sometimes'. One can deducefrom this that some staff members that address gender/WID issues in their operational work, do sohaving never read either OP4.20 or OMS2.12, but having only heard about them. Gender, as in all ofthe questions above, is not related to the answers. Years in Bank, task team leaders and managers areall positively related to addressing gender/WID issues in work.

1In order to target non-administrative Bank staff members in each region, OED used the following Lotus Notesdistribution lists: Afrrlt, Afrtls, Afrttl, Lacrmg, Fullrmt, Lcrtm, Eca 18-21 Hq, Eca Ptls, Eca Sm, Ecasd, Ecacd, Eca Rlt,Mnahl, Sasrmt, Sas Region Task Leaders, Sar Task Leaders, Sased Team Leaders, Sasen Team Leaders, Saseg TeamLeaders, Sasfp Team Leaders, Sashp Team Leaders, Sasin Team Leaders, Saspr Team Leaders, Sasrd Team Leaders,Sassd Team Leaders, Eapregionhlfield, Eaccfhl, Eapregionhl, Eapregionhlhq, Eacifhl, Eackfhl, Eacpfhl, Eactfhil, Eacvfhl.

57 Annex VII

"If you do not integrate gender considerations in your work, please explain why not. " "Lack orresources" and "lack of tools" made up 56 percent of responses. 19 percent answered "genderconsiderations are not relevant because a well designed project will take care of these issues." 13percent answered "lack of time". Ten percent were not aware of the policy. And only three people (2percent) answered "this is a feminist cause not related to development effectiveness."

"What additional assistance would you requirefrom the Bank to improve you work in addressinggender issues?"52 percent answered "additional resources/time". "Clearer guidelines" and "training" both received 9percent of responses while 30 percent answered "all of the above".

2. Finally we asked respondents to choose from among the following objectives, up to threewhich they believe to be relevant for poverty reduction. These results are:

Which of the following objectives are relevant for effective poverty reduction?

Objectives Frequency Percent

* Improving education for girls 366 94* Women's reproductive health 198 51* Ensuring that poor women have

access to financial services 193 49* Addressing relationship between

gender & economic infrastructure 77 20* Violence against women 74 19* gender and land ownership 72 18* Ensuring that poor women farmers

have adequate information andaccess to agricultural inputs 65 17

* Eliminating gender stereotypingin science and technology 34 9

* Identifying and addressing genderissues in private sector 28 7

58 Annex VII

3. The survey also allowed for written comments from respondents. Many people took the time todo so. The following box summarizes the main points of these comments.

Summary of Staff Survey Comments

* To hold senior management accountable for progress on gender with monitoringindicators. Many managers are not convinced this is an important issue. Those that doa good job on gender should be rewarded.* To not force the gender issue in every sector area which leads to a fragmentationof the Bank's work and loss of focus. Forcing the gender issue may buy some votes insome political circles but will not improve the effectiveness of the Bank's projects.* To recognize that gender is not relevant in some sectors, as in the case of energy.* To provide staff with resources to implement gender-sensitive project designs withgender monitoring.* To provide gender training for Bank staff and for clients. Staff need a model, aproduction function, that shows the inputs which are needed to reduce gender-basedpoverty and what is the relative contribution of each input to the desired outcome.* To stress the message that the issue of gender is simple (in an operational sense)and dealing with it is the right thing to do.* To promote gender in local languages.* To not make gender mandatory, as with environment: it creates resentment bystaff. It leads to sterile filling-in of boxes in standard documents and has very littleimpact.* To include men gender and development (i.e. educational achievement of boys;violence by men against men; poor men and access to financials services; malealcoholism, etc).

59 Annex VIII

ANNEX VIII: PROJECT APPRAISAL DOCUMENTS REVIEW

1. OED's Evaluation 2000 did an independent verification of the PREM WID Ratings sytem andestablished benchmarks for integration of WID/gender issues into Bank project documents. Thepresent evaluation assessed 53 PADs in eleven countries.'

2. A 'before and after' assessment shows improvements with regard to gender integration in Bankprojects in these eleven countries. The percentage of projects with no mention of women or genderdecreased, while at the same time those with specific activities for women or gender increased. Theevaluation finds a steep rise in the percentage of projects that use participatory approaches. This maybe misleading, however, because PADs now include a section called "participatory approach" andtherefore, task teams need to report on this aspect. SARs do not have a similar section, although theprojects may have had participatory processes.

3. The 2001 evaluation included observations not covered in the previous evaluation. Included inthese are whether women, non-govemnmental organizations (NGO) and civil society were involved inthe project (at any stage). These observations can serve as benchmarks for future evaluations.

Evaluation 2000 PAD Evaluation 2001Project Characteristics (27 countries) (11 countries)

% of projects % of projectsProject that do not mention of womenand/or gender 62 55Projects with activities for women orgender 25 43Projects with WID/gender analysis and/oractivities for women or gender 38 45

Projects with participatory processes 2 22 96Projects which explicitly involve womenin above participatory processes NA 32Projects including sex-disaggregatedmonitoring indicators 20 32

NGOs involved NA 68

Civil society involved NA 89

' This evaluation reviewed the 5 most recent PADS in II countries (Haiti had no PADs). This resulted in 53 PADs.2 The steep rise in number of projects with participatory processes may be due partly to the fact that PADs unlike SARsrequire explicit explanations of the nature of participation in the preparation and design of the project..

60 Annex VIII

Comparison Between 2000 and 2001 Evaluations

100 /80 _ _ _ __ _ 6

40 I&20/ v

4e

""Evaluation 2000 U Evaluation 2001

61 Annex IX

ANNEX IX: WOMEN'S EYES ON THE WORLD BANKEVALUATION OF HUMAN DEVELOPMENT PROJECTS IN

LATIN AMERICA

1. The Women's Eyes on the World Bank Campaign was launched during the 4 th World Conferenceon Women in Beijing, China (1995) in order to monitor World Bank compliance with a series ofcommitments required to fulfill the Action Platform.

2. Towards this end of the campaign, the Latin American region had analyzed five health projects,four education projects, and two social development project carried out in 10 countries in the region:Mexico, El Salvador, Panarna, Colombia, Venezuela, Peru, Bolivia, Chile, Argentina and theDominican Republic. The World Bank did not directly finance any of the projects analyzed but theBank did influence the project design.

3. This monitoring exercise was carried out using a common research methodology and establishedthree general categories of analysis:

(i) Equity, which is important insofar as Bank - impelled health and education reforms arebased on the principle of encouraging equity among social sectors, re-designating resourcesto more vulnerable sectors of the population. In this sense, it is important to also investigatewhether or not such policies promote gender equity;

(ii) Participation, which is based on the fact that the Bank has established policies aboutparticipation of all interested or affected groups in its projects. As such, monitoringeffective compliance with this principle requires assessing whether or not poor women areparticipating in project operations from the design state to project evaluation; and

(iii) Compliance, which is key if we are to analyze congruence between the Bank's discourseand project implementation actions at the local level, which may also involve an analysis ofgovernment compliance with Bank directives. These analytical considerations resulted in amethodology that included a thorough revision of all relative project documentation,including both Bank and Federal and Municipal Government documents. Furthermore, aseries of interviews were conducted with Bank staff involved in project design,functionaries from central and municipal governments, and various affected personsincluding health and education professionals and project beneficiaries and non beneficiaries.

4. The most significant findings of this investigation are:

Equity: While in eight of the 11 projects (in 10 countries) analyzed, World Bank resourcesare indeed being channeled in the direction of the poorest social sectors, in. only 3 of theseprojects are resources being distributed to the poorest women.

Participation: While the World Bank asserts that there must be beneficiary participation intheir projects, in reality this is occurring in only three of the 10 countries studied.Furthermore, even in these three, there were problems noted in the process; i)The concept of

62 Annex IX

"participation" effectively means using the project service, and ii) in no countries wereresources designated specifically for participation even though this is a dynamic that has areal economic cost. Nevertheless, it is important to mention that when an effectiveparticipation process was carried out, observed efficiency and efficacy of the project weresubstantially improved. This was the case in both Bolivia and Peru.

Compliance: In terms of gender there is serious incongruence between what the Bank says itwill do and what it actually does. For example, while in 6 of the 10 countries analyzed agender component was integrated into the project (and was later incorporated in toGovernment projects as well) the concept of gender used was intimately related to maternity.As such, crucial themes, such a teen-age pregnancy, were ignored and not incorporated intospecific projects.

5. In sum, the analytical exercise elaborated by the Campaign concludes that the advances madein Washington were not observed in the application of these 11 projects. Further still, the lack of anadequate and uniform understanding and conceptualization of gender in the Bank results in the factthat gender is not addressed as an equity issue. As such, discrimination against women deepens inour societies while traditional gender roles are further consolidated.

6. Recommendations to the World Bank resulting from this analysis are: (i) The inclusion ofgender equality as a governing principle of heath and education sector reforms that ostensibly searchfor equity; (ii) a theoretical standardization of the concept of gender to be used by all personnel; (iii)implementing affirmative actions aimed at women with the concrete objective of impelling genderequality, including an increase in resources directed towards women and the integration of a genderperspective in all Bank operations; (iv) the inclusion of mechanisms that assure governmentcompliance with World Bank recommendations concerning gender; (v) re-conceptualizing theconcept of participation that includes content beyond that of defining participation as "service user,"and above all; and (vi) that health and education reforms must consider the integral distribution ofservices to the poor, and not just a "minimal level."'

'Participant organizations by country: Mexico: Gender and Environment Network, Gender and Economy Network,Journalists Network - CIMAC, Civil Organizations Network For a Feminist Millennium. El Salvador: Las Dignas.Panama: CEALP. Colombia: Dialogo Mujer. Venezuela: CISFEM. Peru: Structural Adjustment and Women Group.Bolivia: Women Coalition. Chile: Hexagrama Consultants. Argentina: CISCSA. Republica Dorninicana: PRO FAMILIA.

63 Annex X

ANNEX X: GENDER AND ADJUSTMENT IN TEN COUNTRIES

SUMMARY OF EVALUATION

1. Adjustment programs may affect poor women and men differently. In countries with highlevels of gender disparity, women are not able to access from the benefits and opportunities ofeconomic adjustment and sometimes bear a disproportionate share of the cost of adjustment. There isno consensus among development practitioners on whether it is possible to identify ex-ante thegender-disaggregated impact of individual adjustment programs, given that the impact is dependenton a number of other extraneous variables in the country. Furthermore, neither OP 4.20 nor theOperational Memorandum on Adjustment Programs require that individual programs take genderconsiderations into account.

2. OED, therefore, mainly asked the following questions:

* Based on a desk review, to what extent have the results of adjustment programs been adverse forwomen?

* Did the Bank make reasonable attempts to monitor and respond to the gender-differentiatedresults of its adjustment assistance during transition in a timely manner?

* Did the Bank integrate gender considerations into social safety net or in other projects thataccompanied adjustment programs to cushion their initial costs?

3. Fifty-one adjustment operations that were approved in or after FY 1988 and parallel support tosocial protection programs from ten of the evaluation countries were reviewed.' The analysisreviews project documents, analytical work and country assistance strategies. The evaluationconcludes that the Bank did not support effective monitoring systems that would have facilitated theidentification of any adverse impact on women/gender, and permitted appropriate responses in realtime. Bank assistance was also not effective in fully taking into account gender considerations inpreparing safety nets or other social protection measures.

Results of Overall Adjustment Programs

4. Fifty-one adjustment operations that were approved in or after FY 1988 were reviewed in thisstudy. Most adjustment operations in the sample are themselves gender blind. Only, ten mentiongender or women's issues. Out of the five adjustment programs in the sample that affected socialsectors directly, only two refer to women.2 One of five agricultural adjustment operations statesspecifically that women will benefit.3 There are similar specific references to women's issues inappraisal documents of adjustment operations in Industry, Public Sector Management, Water Supplyand Sanitation, and Multisectoral Projects, but these do not go beyond an identification of gender

' The Gambia and Haiti are excluded since they did not have a Bank adjustment operation that closed between FY 95 andFY 00, which is the criteria used in this evaluation.2 These are the Zambia Economic and Social Adjustment Credit (FY 94) and the Cote d'Ivoire Humnan ResourcesAdjustment (FY 92). However, this reference is not always based on sound analysis and does not lead to specific actions.For example, the Zambia operation merely asserts that the changes brought about will benefit women relatively morebecause it will reverse certain discriminatory practices in land tenure practices.3 This is the Philippines Environment and Natural Resource Management (FY 91). At the design stage, the project wasexpected to benefit women through developing rural infrastructure.

64 Annex X

disparities in selected indicators or assertions that women would be beneficiaries from the operation.None of the 51 projects included a meaningful analysis of gender issues.

5. Existing evidence reflects that some adjustment programs supported by the Bank loans hadadverse outcomes for the poor in general. 4 For example, in Kyrgyz, Zambia, Cote d'Ivoire, Ecuadorand Yemen growth rates declined/stagnated and/or poverty increased in the 1 990s.5 On the otherhand, in Poland and Vietnam, the adjustment measures achieved relative success in terms of growth.However, it is difficult to assess the gender-differentiated results of these programs. The evaluationdraws on a body of evidence both internal and external to the Bank to make limited judgments basedon outcomes and trends in key indicators of well-being. While these trends are not meant to indicatecausality, they describe how gender-disparities shifted during the adjustment period.

6. In Poland and Vietnam, given the general success of the transition measures and the low degreeof existing gender disparities, it is likely that men and women benefited to a similar extent from theadjustment programs. In Vietnam, women, especially young, single women, had new opportunitiesto earn wage incomes during the Doi Moi reforms as export growth sectors-particularly textiles andshoes-attracted large number of women from rural areas. However, even in these countries, certainoutcomes like higher childcare costs in Poland and rising costs of services and reduced provision ofstate-run creche facilities that accompanied the transition in Vietnam placed more demands onwomen's time for childcare and care of the elderly. In Yemen, Bangladesh, Philippines and SriLanka adverse effects on women are not clearly discernible. In Kygyz, Zambia, Cote d'Lvoire, andEcuador, the results of the adjustment regime for women is likely to have been negative in severalsectors. In Zambia, for example, in-country consultations suggest that rising costs of education andcommercialization of agriculture were particularly detrimental for females. Though both men andwomen were affected during the transition in the Kyrgyz, the deteriorating education and health caresectors have reduced employment possibilities for women in sectors where they dominated. Reducedsocial assistance for children and the elderly has intensified domestic work burdens. In Coted'Ivoire, microeconomic evidence suggests that public sector downsizing may have affected womendisproportionately as there was less wage discrimination in this sector. In Ecuador, medical care wasoften deferred during developments of the 1990s, and the percentage of births attended by medicalpersonnel declined to 64 percent in 1998, from 84 percent in 1987.

Monitoring of Gender Disaggregated Impact

7. Given the difficulty in anticipating all possible gender-differentiated outcomes of wide-rangingmacroeconomic policies, the evaluation also assessed whether the Bank had instituted effectivemonitoring systems that would have permitted the Bank to track and understand any adverse impacton women, and respond in real time with appropriate measures.

8. Only one of the 51 adjustment operations monitor gender issues within their structure, and noneprovide gender-disaggregated monitoring indicators. 6 Most of the social protection measures werenot monitored for outcomes on men and women separately either, and are broadly poverty-focused.

4 This information is obtained from selected Poverty Assessments and background data in Country Assistance Strategydocuments. Data from WDI indicators were also examined to confirm this.'This is based on evidence in Bank documents and data on per capita incomes.6 This is the Zambia Economic and Social Adjustment Credit (FY 94) which rates the achievement of gender concerns as'Negligible' in the ICR.

65 Annex X

Bank studies for the Kyrgyz Republic, Zambia, and Poland show some evidence of monitoringoutcomes for men and women at a macroeconomic level. The Bank's 1994 Poverty Assessment forPoland has a whole chapter devoted to women's welfare during the transition. There is evidence of aslight upturn in maternal mortality rates during the transition in the Kyrgyz Republic, which isrecognized by the World Bank-but there are no other instances of monitoring. Overall, themonitoring of gender disparities during transition is unsatisfactory (Table 1).

Integration of Gender Issues into Accompanying Operations

9. Table 1 below indicates that there were accompanying safety net or relevant operations in allcountries, but in general these did not attempt to mitigate adverse impact on women specifically.Some examples are provided below.

Table 1: Integration and Monitoring of Gender Issues during Adjustment

Countrv Severity of Gender Integration of Gender into Monitoring of Gender Issues in________ Disparities' Social Protection Measures' Adjustment Measures3

Kyrgyz M N NEcuador M M NZambia H M MBangladesh H M NVietnam M M NYemen H M NPhilippines M N NPoland N M MSriLanka M M NCote d'lvoire H M NAverage SU M NNote: All ratings are on a four-point scale, where N = Negligible, M Modest, SU = Substantial, H = High.

1. This is based on a relative ranking of countries based on the UN Gender-related Development Index.2. This rating is provided by the evaluation based on an integration of gender issues into the design of social protection measures that are

financed by the Bank. These measures are not only the ones classified by the Bank as 'Social Protection', but also include social protectioncomponents in other Bank-supported projects of the 1990s.

3. This rating is provided by the evaluation based on monitoring of gender-disaggregated outcomes in adjustment measures supported by theBank in the 1990s.

10. In Zambia and Cote d'Ivoire, given unfavorable poverty outcomes in the 1990s, severe genderdisparities, and the importance of the adjustment program, greater attention to gender in the contextof adjustment might have been warranted. In Zambia, the World Bank had two Social RecoveryProjects (FY 91 and FY 95). While there is some analysis of gender disparities, the initial design ofshe operations did not include any specific measures to ensure that men and women would benefitequitably.7 For example, the projects included provision of temporary employment from communitysub-projects from which mainly men benefited. On the other hand, communities needed tocontribute free labor to the sub-projects, which was provided mostly by women. In Cote d'Ivoire,specific social protection measures included the Labor Training Support Project (FY 94) that waspart of a Vocational Development Training Fund program. The Labor Training Support Project andincluded training and apprenticeship measures for informal sector workers. This project envisaged

7 The Second Social Recovery Project did incorporate measures to ensure women's participation as it progressed, largelydue to the initiative of the Project Management Unit. However, the ICR rates the achievement of gender objectives as'Modest'.

66 Annex X

women as a primary group of beneficiaries as they were concentrated in the informal sector, which isthe target for the project, but there was no clear action to ensure that this happened.

11. Evidence suggests that men and women benefited similarly from adjustment programs inVietnam and Poland. Both countries are characterized by the absence of significant genderdisparities, and had large-scale adjustment programs in the 1 990s, which saw impressive growth anddeclining poverty. In Vietnam, one of the key elements of the country assistance strategy was toprotect the poor from the adverse impact of the adjustment. While there were number of investmentprojects in this regard, none of them took into account gender considerations in design, even thoughLSMS Surveys that were conducted with Bank assistance provided the information base. In Poland,much of the social protection was gender blind as well. Only the Employment Promotion andServices Project (FY 91) that sought to provide support to mass layoffs associated with adjustmentwas based on some analysis of gender issues on the labor market, and recognized women as avulnerable group. In neither of these countries do we see an adequate gender-sensitive strategy ofsocial protection. However, given country conditions, this was not a major drawback of the Bank'sprogram.

67 Annex XI

ANNEX XI: REPORT FROM CODE

Committee on Development Effectiveness(CODE2001-0054, May 29, 2001)

Integrating Gender into the World Bank's Work: A Strategyfor Action (Discussion Draft)and The Gender Dimension of Bank Assistance: An Evaluation of Bank Results

1. The Committee met on May 9, 2001 to give feedback to Management on the Gender MainstreamingStrategy Paper entitled Integrating Gender into the World Bank's Work: A Strategy for Action (DiscussionDraft) (CODE2001-0043) prior to Board discussion of the strategy, and to review the findings of the OEDstudy entitled The Gender Dimension of Bank Assistance: An Evaluation of Bank Results (CODE200 1-0044). OED said that their report was the culmination of two years of work on evaluating the genderdimensions of Bank assistance. The work on the evaluation had occurred in close coordination with theGender and Development Board and the findings had been incorporated on an ongoing basis into thepreparation of the Bank's strategy. OED summarized the main findings of the evaluation and reiterated itsrecommendations which suggested that the Bank clarify its gender policy, including its implementation; basethe proposed strategy on comprehensive diagnosis of country gender contexts; explicitly integrate genderconsiderations into CASs; mainstream gender into economic and social analysis; and establish a system tomonitor policy implementation and progress on the ground. OED noted that, overall, the draft strategyaddressed these recommendations, but that there were some remaining concerns. In particular, OED stressedthat the draft strategy went beyond the scope of OP4.20 and thus the policy required revision in the nearestfuture to be consistent with the proposed strategy and that a BP was required to give clear instruction to Staff.OED urged Management to set a clear timetable for this exercise. OED also emphasized the importance ofcountry ownership of the new approaches set out in the draft strategy and the need to consult with clientsbefore finalizing it. Lastly, OED urged Management to establish realistic, phased, and monitorable results-oriented benchmarks to evaluate implementation.

2. Management welcomed the discussion and thanked Executive Directors and OED for their input to date.Management noted that the draft strategy provided a broad road map for integrating gender into the Bank'swork with the intent of reducing poverty and enhancing economic growth. Management stressed that the draftstrategy proposed a process for gender analysis and action on a country-by-country basis recognizing thatgender conditions were distinct in each country. Management also emphasized that the draft strategy proposedthat the Bank's consultations be gender inclusive and that the Bank take a proactive but supportive role in itsdialogue with Governments in raising gender issues on the basis of the country-by-country analysis. The draftstrategy also suggested organizational changes to facilitate implementation, including clarifying roles andresponsibilities, accountabilities, training needs, and financing. Management agreed with OED that strongmonitoring and evaluation systems were critically important and noted that it was a priority of the Gender andDevelopment Board to work with the Regions to develop these systems. Management proposed that theGender and Development Board report annually to EDs on the status of gender mainstreaming in the Bank.

3. The Committee welcomed both papers and commended Management and OED for well-written andthoughtful documents. The Committee considered the draft strategy responsive to the OED evaluation andparticularly welcomed the accountability matrix as good practice. The Committee encouraged Management toaccelerate the implementation of the draft strategy through more ambitious, time-bound action plans. Amongthe specific issues raised by the Committee were:

68 Annex XI

4. Leadership and Accountability. Members stressed the need for corporate accountability and committedleadership and asked that the draft strategy be refined to describe concrete actions Senior Management wouldtake to operationalize it. Some also emphasized the need for the Board to play a leadership role in this areaand suggested a training seminar on gender issues be held for Board members. Members emphasized thatRegional VPs and their Staff needed to be held more accountable for developing and implementing RegionalGender Action Plans and noted that the annual reporting should be done by the VPUs rather than by theGender Board. They noted the need to move forward quickly with the Regional Action Plans and askedManagement for a timetable in the revised draft strategy.

5. Institutional Capacity. The Committee strongly supported the OED recommendation that Borrowerpolicies and institutional capacity to address gender issues be strengthened, and noted that the draft strategywas not sufficiently responsive to the recommendation. They also noted that capacity building should be apart of the Bank's support to Borrowers and should not be done only on a cost-recovery basis. The Committeeasked Management to strengthen this aspect in the revised draft strategy.

6. Gender OP/BP. The Committee strongly supported the need to revise the Gender OP to reflect the draftstrategy, and the need for a BP that clearly outlined procedures for Staff to follow. They asked Managementto clarify the timeline for OP/BP preparation in the draft strategy and stressed that this process should becompleted in the nearest term.

7. Monitoring and Evaluation. The Committee noted the draft strategy's relatively weak treatment ofmonitoring and evaluation and emphasized that this needed strengthening and that inclusion of measurementof impact in Borrowing countries would be advisable. Management agreed with the importance of monitoringand evaluation and noted that the Gender Board was working with the Regions in this area.

8. Link to PRSP. The Committee asked that the draft strategy be more explicit about how the Bank willensure that gender analysis done in each country has an impact on outcomes. In this regard, memberswelcomed the proposed links to the JSA and PRSP noting this was an important point of entry for genderanalysis and dialogue with Governments. One member stressed that it was a sequencing issue and it was moreimportant to first promote gender awareness in Borrowers or risk jeopardizing the country ownership ofPRSPs.

9. Funding. Members urged the Bank to allocate sufficient resources for the implementation of the strategyand asked to see the details of the Bank's financial commitment to the draft Gender strategy in the upcomingbudget proposal for FY02. Noting resource constraints, some members asked why full-time GenderCoordinators were needed for each Region and stressed that providing training for Country Managers and theirteams would be more efficient and cost-effective. Management confirmed that its commitments supportingthe strategy would be highlighted in the budget for FY02.

10. Cost-Benefit. One member questioned whether a cost-benefit analysis had been done. In particular, heasked about the costs to Borrowers for implementing the draft strategy. Another member wondered abouttradeoffs and stressed the need for prioritization.

11. Adjustment Lending. There was discussion about the lack of attention to gender in adjustmentoperations, and members stressed the importance of gender analysis in the Bank's adjustment lending. Theyalso noted that the OED evaluation had not paid adequate attention to this issue and asked for clarification.OED responded that a separate study had been done on this issue but space limitations had made it difficult toinclude all their-findings in the evaluation. A working paper on the findings would be published separately.Management clarified that all issues related to adjustment lending policy would be taken up in the context ofthe conversion of OD 8.60.

69 Annex XI

12. Partnerships. The Committee stressed that many other agencies had more experience in the gender areaand urged Management to work with them, particularly citing the UN. Speakers also asked for a greaterinclusion of IFC in the Gender Mainstreaming Strategy noting that OED's findings had pointed to privatesector development as an area of weakness.

13. Next Steps. Management agreed to provide written responses to the specific questions raised as soon aspossible. Management will also proceed with external consultations and revise the draft strategy taking theCommittee's comments into account. It was hoped that the draft strategy would be ready for Board discussionin September 2001 or at the earliest date thereafter based on availability on the Board schedule. TheCommittee recommended that the Board approve public disclosure of the OED evaluation during its approvalof the draft strategy.

Pieter Stek, Chairman

DistributionExecutive Directors and AlternatesPresidentBank Group Senior ManagementVice Presidents, Bank, IFC and MIGA

70 Annex XII

ANNEX XII: OED PRESENTATION AT THE CODE MEETING

Chairperson, Members of CODE:

This report represents the culmination of two years of work in OED on evaluating the genderdimensions of Bank assistance. We worked in close consultation with the Gender and DevelopmentBoard and transmitted our findings on an ongoing basis into the preparation of the Bank's strategy.Last October we presented to you a report focused on the integration of gender into Bank Assistanceand previously in March 2000, we issued the proceedings of the Entry Workshops for this evaluation.Over the next year, OED plans to publish: (i) a volume with twelve country gender studies, (ii) theproceedings of four regional gender workshops held in Manila, Nairobi, Quito and Warsaw, (iii) thebeneficiary assessments undertaken for this evaluation; and (iv) a working paper on the results ofstructural adjustments in 12 sample countries; (v) a Primer on Lessons for Integrating GenderConsiderations into Bank Operations. Today, we would like to summarize the main findings of ourevaluation, reiterate our 6 recommendations, and present our opinion on whether the proposed genderstrategy addresses OED's recommendations.

First, our main findings. Last year, we reported that 70 percent of the CASs between FY97 andFY00 integrated gender considerations, a much higher number than those before 1997. We find thatthere is satisfactory gender mainstreaming into the health and education sectors, and we are indeedpleased to confirm that in these sectors the results on the ground have also been positive.

We find that Bank assistance is most effective when five elements are present: it is based on acomprehensive diagnosis of the gender context, the assistance is not sectoral but morecomprehensive, framed within a country-led agenda, delivered with strong country ownership atdifferent levels, and in strong collaboration with other partners. However, in our sample of 12countries this was so only in Gambia and Bangladesh.

We also find that outside the health and education sectors, integration into Bank assistance in theeconomic sectors has been weak and ad hoc. Whether it is the financial sector, the public sector,small scale and medium enterprises or the private sector, gender is mostly absent or only marginallydiscussed. This is true at the global strategy levels, at the analytical levels, and at the project levels.Bank assistance has thus fallen short in enhancing the economic participation of women.

The present evaluation also confirms what we told you earlier: the gender policy framework isstill unclear; the responsibilities and processes for gender have not been established; and systems formonitoring results and policy implementation are weak.

What does OED recommend?

OED's main recommendations are that the Bank clarify its gender policy and take into accountcountry policies for women or gender and strengthen local institutions to implement these policies.We propose that the Bank clarify the roles and responsibilities for implementing the gender policywithin the Bank. We also suggest that any proposed strategy be based on comprehensive diagnosisof country gender contexts, explicitly integrate gender considerations into Country Assistance

71 Annex XII

Strategies, and mainstream gender into the economic and social analysis that is already beingundertaken during project preparation. In addition, we seek that the Bank establish a system tomonitor policy implementation and progress on the ground.

Now let me turn to the question of whether the proposed strategy addresses OED'srecommendations?

By and large, it does. It clarifies the responsibilities and suggests processes for gendermainstreaming. It proposes the development and implementation of a monitoring system that willfocus on results on the ground. It proposes that Bank assistance be provided within a country ledagenda. It supports comprehensive analytical work that will identify issues for integration intoCASs, other products, and as necessary into projects. We continue to have some concerns.

First, the proposed strategy seems to deepen and widen the present gender policy. The proposedgender strategy seems to be based on a broad analytical framework of gender equality that goesbeyond the poverty reduction objectives of the existing policy, which focuses on enhancing theeconomic participation of women. OP 4.20, therefore, now needs not only to be clarified as we hadrecommended last October, but also to be revised consistent with the proposed strategy. In addition,the strategy provides for such operational flexibility that OED is not clear on what is actuallyrequired and what is optional. OED therefore stresses the need for a Bank Procedure (a BP) toinstruct staff on their responsibilities and establish accountability and consistency in procedures. TheGSP indicates the need for a revised OP and BP, but OED urges Management to set a clear timetablefor this.

Second, although the arguments for the shift to gender equality as discussed in the GSP arecompelling, we note that the policies of many clients still stay within the paradigm established by the1995 Platform for Action for Women. Clients, therefore, continue to focus on the advancement ofwomen. It is, therefore, important that the Bank consult with, and convince, clients of this newapproach before adopting the revised policy. Greater ownership of this policy outside the Bank isalso required if it is to be effectively implemented in client countries. Otherwise, the Bank runs therisk of being perceived as imposing externally driven agendas.

Third, the evaluation finds that the Bank achieved limited results, although its present objectivesare far less ambitious. The proposed strategy plans to address priority gender issues in all clientcountries. The Bank once again faces a vulnerable situation of promising too much and risking thepossibility of achieving little. To avoid this, OED suggests that Management establish realistic,phased, and monitorable results-oriented benchmarks for policy evaluation and simultaneouslyencourage partnerships to allocate responsibilities for achieving institutional goals. Otherwise,another evaluation five years from now will likely judge the Bank harshly, only because it haspromised to do what it may not be able to achieve.

Finally, as a woman from a sub-continent teeming with poverty and gender issues, permit me toend on a personal note. Over the last two and half years, I have had the privilege to hear fromwomen and men from many parts of the world. Very few disagree on the need for gender justice,although most reiterate that the paradigm for change must come from within the country. This canhappen only if both women and men can participate in development decision-making and equitably

72 Annex XII

enjoy the benefits of development. The Bank's biggest contribution and role will lie in strengtheningprocesses and mechanisms for the equitable participation of both women and men. We hope thatsuch participation will also be a strong pillar of the Bank's future gender policy and strategy. Let meclose with the voices of African women from the Nairobi workshop: "Do not treat us as powerlessand vulnerable victims. Empower us with resources, human capital, and assets. We will change ourgender relationships."

73 Annex XIII

ANNEX XIII: EXECUTIVE SUMMARY OF THE GENDERSECTOR STRATEGY PAPER

Executive Summary

Several major World Bank reports provide strong empirical evidence that the gender-baseddivision of labor and the inequalities to which it gives rise tend to slow development, economicgrowth, and poverty reduction.' Gender inequalities often lower the productivity of labor, in both theshort term and the long term, and create inefficiencies in labor allocation in households and thegeneral economy. They also contribute to poverty and reduce human well-being. These findingsmake clear that gender issues are an important dimension of the fight against poverty. Gender issuesare also central to the commitments made by the World Bank's member countries at the UnitedNations Millennium Summit and at the Fourth World Conference on Women in Beijing in 1995.Although the nature and importance of gender issues for poverty reduction and growth vary fromcountry to country, significant gender disparities are found in all regions of the world (including inthe member countries of the Organisation for Economic Co-operation and Development-OECD).These disparities tend to be greater in low-income than in higher-income countries, and, withincountries, tend to be greater among the poor than the more affluent. The incorporation of genderissues into development actions needs to be sensitive to the specific conditions in each country.

Since the 1980s, the Bank has made progress in integrating gender issues into country workand lending, particularly in education and health. For example, between 1995 and 2000 the Bank lentmore than $3.4 billion for girls' education programs, and was also the single largest lender in theworld for health, nutrition, and population projects, three-quarters of which contained gender-responsive actions. Attention to gender issues in World Bank Country Assistance Strategies (CASs)also increased during this period. Several organizational changes designed to facilitate greaterattention to gender and development issues were also made, including issuing of an OperationalPolicy on the gender dimension of development in 1994, and creating a Gender and DevelopmentBoard and placing it within the Poverty Reduction and Economic Management Network in 1997.

Despite the progress to date in gender mainstreaming, the Bank's effectiveness can beimproved by paying more systematic and widespread attention to gender issues in the context of ourpoverty reduction mandate. The opportunities for improving the development impact of the Bank'swork through gender mainstreaming include making Bank interventions more responsive to countrygender conditions and commitments; making these interventions more strategic; and improving thealignment of Bank policies, processes, and resources to support such interventions.

In the strategy described in this paper, the World Bank will work with governments andcivil society in client countries, and with other donors, to diagnose the gender-related barriersto and opportunities for poverty reduction and sustainable development; and will then identifyand support appropriate actions to reduce these barriers and capitalize on the opportunities.

Engendering Development-Through Gender Equality in Rights, Resources, and Voice, 2001a; WorldDevelopmentReport 2000/2001: Attacking Poverty, 2000d; and Voices of the Poor: Can Anyone Hear Us?, Narayan et al, 2000.

74 Annex XIII

The strategy is intended to establish an enabling environment that will foster country-led,country-specific strategies for changing the gender patterns that are costly to growth, povertyreduction, and human well-being. The strategy rests on a basic process that involves working withcountries to:

* prepare periodic, multi-sectoral Country Gender Assessments (CGAs) that analyze the genderdimensions of development across sectors and identify gender-responsive actions important forpoverty reduction, economic growth, human well-being, and development effectiveness, andwhich inform the Bank's country assistance program;

* develop and implement, as part of the Bank's country assistance program, priority policy andoperational interventions that respond to the assessment; and

* monitor the implementation and results of these policy and operational interventions.

A key component of the strategy is the CGA, a country-level gender analysis that identifiescritical areas in which gender-responsive actions are likely to enhance growth, poverty reduction, andwell-being in a particular country context. Country Directors will ensure the completion of theseassessments in a timely manner in countries with an active lending program (and in which anassessment has yet to be conducted). In line with the ongoing reform of the Bank's analytical work,the methods for carrying out these assessments will be flexible. The country gender analysis may, forexample, be a stand-alone document or a section of a country poverty or economic analysis. TheCGA may contain original, analytical work or may simply refer to such work produced by the Bankor by other agencies (government, international, academic). To lower costs, increase buy-in, andbuild on expertise outside the World Bank, collaboration with government, civil society, and otherdonors in completing the CGA will be emphasized when possible. Management, with the assistanceof the Gender and Development Board, will clarify standards for conducting the assessments andmonitoring their quality. Updates will be done in accordance with the typical cycle for countryanalytical work (approximately once every five years), and will be designed to investigate whetherany major changes in gender conditions have occurred (i.e., updates will typically be less elaboratethan the initial assessment). When possible, CGAs that contain substantial original work will bepublished and shared broadly.

The gender strategy rests on four actions designed to enable the process of diagnosis, strategyformation, and integration into operations:

Integrating a gender dimension into relevant analytical work and lending instruments. Anunderstanding of gender issues in sectors identified in the diagnosis as important for povertyreduction and growth is a necessary prelude to gender-responsive project design. Tracking whetheranalytical work and projects in these sectors have been conducted with a cognizance of key genderissues is also important for improving performance and quality. For this reason, the strategyrecommends integrating a gender dimension into:

* sectoral analytical work in high-priority sectors (as identified in the CGA), and into thesocial impact analysis associated with adjustment lending (including Poverty ReductionSupport Credits);

75 Annex XIII

* the criteria used to assess the poverty reduction focus of Poverty Reduction StrategyPapers (PRSPs) in Joint Staff Assessments (JSAs);

* the criteria used to assess the adequacy of the poverty reduction focus of the CAS andSector Strategy Papers; and

* the quality criteria used by the Quality Assurance Group to assess projects and analyticalwork, with appropriate consideration of whether particular sectors have been identified ashigh priority in the CGA.

Supporting the strategic integration of gender issues into operations. The strategyrecommends three forms of support for operations:

* Training. To assist Bank staff and counterparts in strategically mainstreaming genderissues into operations, the Gender and Development Board will work with HumanResources and the World Bank Institute to add gender and development content to theBank's staff orientation course and to existing core courses. Selected technical modulesoriented to operational staff will also be developed.

* Operational tools. Together with regional sectoral staff, the Gender and DevelopmentBoard will continue to create and disseminate adaptable tools and good practice examplesfor use in operations. Good practice examples, including those involving macroeconomicpolicy advice, will be emphasized.

* Building capacity in implementing agencies. The Gender and Development Board willexplore methods to provide technical advice to implementing agencies. The Board willalso work with the regions and the World Bank Institute to identify and support trainingopportunities for clients and counterparts.

Aligning resources with the elements of the strategy. Four types of resources will be alignedwith the strategy:

* Budget. In line with the decision to make gender mainstreaming one of the Bank'sCorporate Advocacy Priorities, Bank budget is being redeployed to support genderanalysis and mainstreaming.

* Accountabilities. The responsibilities of Bank staff for gender mainstreaming are beingclarified (Table 2.1). Regional Vice Presidents will ultimately be accountable for regionalresults; they will in turn hold Country Directors and Sector Directors/Managersaccountable for ensuring appropriate integration of gender issues into country operations.To clarify policies and procedures, a revised Operational Policy and Bank Proceduresstatement on gender and development will be issued after due consultation with externalstakeholders and consideration by the Executive Board. While these consultations areongoing, Management will issue an Operational Memorandum to clarify the existinggender policy and provide interim guidance on implementation.

* Staff. Regions will provide in-house technical expertise in gender and development toassist staff in gender analysis and strategic operational mainstreaming, especially duringthe initial years of implementation.

76 Annex XIII

Partnerships. Because of the valuable resources that other organizations have to offer andthe potential synergies with Bank-led activities, the strategy also encourages the formation ofcountry-level partnerships with governments, civil society, and other donors, especially in thecontext of particular projects or programs, including formulation of the PRSPs.

Monitoring and evaluation. Finally, in order to track progress and enhance learning andquality, an effective system of monitoring and evaluation that includes assessment of on-the-groundresults is under development. Progress in implementing the strategy will be reported annually.

The estimated incremental costs of implementing the strategy will be about $2 million in thecurrent fiscal year, about $3 million per year in the three subsequent years, and approximately $2.5million per year thereafter. Corporate incentive funding of $0.6 million has been set side for thecurrent year, and the regions have committed more than twice this amount as matching Bank budget.Implementation has thus been frilly funded for the current fiscal year.

77 ATTACHMENT 1

ATTACHMENT 1: RECOMMENDATIONS OF OEDEVALUATION 2000

1. This evaluation found that staff are largely unaware of the Bank's gender policy, that thereis no consensus on its scope or elements. Further, the Bank has not made adequateimplementation arrangements or established accountabilities for policy implementation. Theevaluation also found that there are no effective systems to monitor and evaluate progressachieved in policy implementation and effectiveness. OED recommends that the followingactions be taken.

Recommendations

2. Clarify the scope of the gender policy. As proposed in the Gender Strategy ConceptNote, the Bank should clarify the rationale, intent, and scope of its gender policy. GSB shouldmonitor the conversion of OMS 2.20, and ensure that provisions for screening projects forimpact on women are revised and included in the new operational policy on socialassessments. Management should require staff to integrate gender considerations into theimplementation of existing social and environmental safeguard policies. Consensus buildingand client consultation should be an integral part of policy formulation.

3. Strengthen management of the gender program. Commitment to gender should bedemonstrated by allocation of resources and training of staff and managers. GSB should beprovided with the required authority to ensure that all networks and families take steps tomainstream gender as appropriate. Each region should establish gender action plans for thenext 36 months, and propose time-bound results indicators.

4. Establish a monitoring and evaluation system. The PREM network, working with theregions, should ensure that institution-wide progress is regularly tracked and periodicallyevaluated. To facilitate monitoring, the existing WID rating system should be improved tomeasure the integration of gender in Bank assistance. Most importantly, the Bank shouldsystematically measure the impact of its assistance to facilitate organizational learning andeffective policy implementation.

5. OED also suggests the following actions:

* The Bank should ensure that all its interventions take appropriate account of genderconsiderations. In particular, gender assessments should become the norm rather than theexception. Individual country programs should make selective use of lending and non-lending instruments, reflecting the intensity of borrower ownership and an explicitdivision of labor with other partners.

* Responsibilities for gender integration should be decentralized from self-standing units tooperational task teams only when there is requisite capacity at the operational level.

78 ATTACHMENT 1

* PCD/PADs should contain a section that: (i) explains whether gender issues wereconsidered and addressed, and if not why not; (ii) identifies the indicators that will beused to monitor the gender-disaggregated impact of the operation; and (iii) establishes asystem through which these indicators will be monitored.

* Gender assessments at the country level should be required as a basis for integration ofgender considerations into country programming and policy dialogue, and as a tool for ofmainstreaming gender considerations into the Bank's assistance.

Examples of Good Practices in Gender

The Tanzania Population, Health, and Nutrition Sector Review of January 1989 made appropriateconmections to the distribution of labor between men and women. It recommended an information,education, and communication program to change attitudes about gender roles, and suggested that theprogram aim toward having men increasingly share the burden of family chores. It also recommendedinterventions to reduce the amount of time and energy women spent on burdensome household chores.

The 1997 Romania CAS contained gender analysis on key gender issues, providing both qualitative andquantitative information. It discussed gender and land issues and specifically included participatorysurveys of land holdings. It planned a special pilot to help women entrepreneurs deal with gender-basedconstraints, and proposed a high-level briefing to increase gender awareness among senior policy-makers.Although the matrix did not include gender-disaggregated monitoring indicators, it referred to a genderaction plan to be included in the Social Development Fund project.

Lessons from a client on allocating resources for gender work. The Philippine govermnent hasrequired by law that each govemment department set aside five percent of its administrative budget forgender mainstreaming. According to the National Commission on the Role of Filipino Women, the lawhas catalyzed gender mainstreaming and has resulted in even departments like Customs and Excise nowturning to them for guidance.

Gender assessments have recently been conducted in Ghana (FY99), Argentina (FY98) and Ethiopia(FY98). These were all undertaken in a highly participatory manner and have been useful in: (i) obtaininga holistic understanding of gender issues in the countries; (ii) providing input for designing more gender-aware products or constituting the basis for more detailed research; (iii) helping understand and establishcontact with the key players, and initiating and establishing dialogue with different stakeholders in thecountries; and (iv) enhancing country ownership and creating demand for Bank-financed products.

The India Uttar Pradesh Sodic Land Project was noted for its participatory preparation andimplementation. It has successfully reclaimed 60,000 ha of sodic soils through efforts of localconmmnities mobilized into self-help groups, including 2,609 Women's Self Help Groups. The projectensured that spouses were registered as co-title holders of allotted lands. It defined a role for women inreclamation through participation in project-specific user groups, notably Site Implementation Committees.It empowered women to play the role of technology transfer agents and animators in a number of areassuch as literacy, health, and animal husbandry.

79 ATTACHMENT 2

ATTACHMENT 2: MANAGEMENT RESPONSE TO THERECOMMENDATIONS OF THE OED GENDER REVIEWS

Major monitorable OED recommendations Management responserequiring a response

1. Clarify the scope of the gender policy. As Management agrees that there is a need forproposed in the Gender Strategy Concept Note, the clarifying the scope of the gender policy. TheBank should clarify the rationale, intent, and scope proposed strategy recommends revising OP 4.20of its gender policy. [The Gender and and writing an accompanying BP as part ofDevelopment Board] should monitor the strategy implementation (paragraph 2.20). Theconversion of OMS 2.20, and ensure that Gender and Development Board, with adviceprovisions for screening projects for impact on from LEG and OPCS, will draft a revised OP/BPwomen are revised and included in the new by 10/30/01. Internal and extemal consultationsoperational policy on social assessments. will be held, and the final draft will be sent to theManagement should require staff to integrate Board of Executive Directors by a target date ofgender considerations into the implementation of 06/30/02. Gender considerations will also beexisting social and environmental safeguard integrated into other operational policies,policies. Consensus building and client including those arising from the conversion ofconsultations should be an integral part of policy OMSs (paragraph 2.20, Table 6.1).formulation.

Management agrees with the general thrust of this

program. Comanitment to gender should be recommendation. However, Management believesprmogstram.dby Comm aitn to gr rsouldsbe that the authority to ensure this mainstreaming

training of staff and managers. [The Gender and should follow the normal lines of authority withintrainipmng t staff ] an old m a g r. h Gder and the World Bank, which are summarized in TableDevelopment Board] should be provided with the 2.1. The proposed strategy lists ensuringrequired authority to ensure that all networks and implementation of key strategy elements, such asfamrilies take steps to mainstream gender as CGAs and appropriate follow-up actions, as a keyappropriate. Each region should establish gender RVP responsibility (paragraph 2.21). RVPs willaction plans for the next 36 months, and propose also be required to submit annual action planstime-bound results indicators. (paragraph 2.21).

3. Establish a monitoring and evaluation system. Management agrees that there is a need for anThe PREM network, working with the regions, adequate monitoring and evaluation system. Thisshould ensure that institution-wide progress is is a key element outlined in the proposed strategyregularly tracked and periodically evaluated. To (paragraphs 2.28-2.29). RVPs will submitfacilitate monitoring, the existing WID rating progress reports annually to the MDs, which thesystem should be improved to measure the GAD Board will consolidate into a Bank-wideintegration of gender in Bank assistance. Most summary to be submitted along with the RVP

' World Bank. 2000. "Integrating Gender in World Bank Assistance." Operations Evaluation Department. Washington,DC. World Bank. 2001b. "The Gender Dimension of Bank Assistance: An Evaluation of Results." Operations EvaluationDepartment. Washington, DC.

80 ATTACHMENT 2

Major monitorable OED recommendations Management responserequiring a response

importantly, the Bank should systematically reports to the Management Committee and, inmeasure the impact of its assistance to facilitate turn, to the Board of Executive Directorsorganizational learning and effective policy (paragraph 2.29).implementation.

4. Strengthen borrower institutions and policies. Management agrees with the thrust of thisBefore and since [the Fourth World Conference on recommendation. Gender training and buildingWomen held in] Beijing, most borrowers have, at capacity in implementing agencies are twothe behest of UN Agencies, fornulated national elements of the proposed strategy (paragraphspolicies and/or action plans for gender equality 2.14-2.16 and 2.28). Using country gender actionand/or the advancement of women, consistent with plans as entry points for country dialogue is alsothe principles of the Beijing Platform for [Action]. recommended (paragraph 1.19), and these plansThe Bank should strengthen development are part of the recommended content of theinstitutions to support the implementation of these Country Gender Assessment (Annex F).policies/action plans. In countries where suchpolicies and plans are weak, support for theirstrengthening (through country dialogue and non-lending services) should be a Bank priority.

5. Integrate gender considerations into country Management agrees that gender considerationsassistance strategies. Based on a comprehensive need to be adequately integrated into CASs. Thediagnosis, the CAS should explain how Bank strategy states that CASs need to refer to existingassistance will take into account the linkages CGAs (paragraph 2.11, third bullet). Managementbetween poverty and gender. This assistance and will take responsibility for ensuring that theits underlying strategy should be explicitly related treatment of gender issues in the CAS isto the Borrower's policy framework for gender. appropriate; every 18 months, the CAS

retrospective review monitors the treatment ofgender issues in the most recent cohort of CASs.

6. Integrate gender into the design of Bank- While Management agrees with the thrust of thissupported projects. Gender analysis should be recommendation, it does not regard the concept ofintegrated into the economic and social analysis "gender disparities" as well defined when used tocarried out in the preparation and design of Bank- characterize whole countries. As part of thesupported projects so that both men and women are Bank's overall country program, the proposedable to access the benefits equitably. This is strategy specifies gender analysis and gender-especially critical for countries with high gender responsive project design in sectors that thedisparities. Country Gender Assessment has identified as

important for poverty reduction (paragraph 2.9).